ARTICLE IN PRESS

Journal of Environmental Management 82 (2007) 24–31

www.elsevier.com/locate/jenvman

Aerobic treatment of palm oil mill effluent

K. Vijayaraghavan, Desa Ahmad, Mohd Ezani Bin Abdul Aziz
Department of Biological & Agricultural Engineering, Faculty of Engineering, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
Received 7 February 2005; received in revised form 28 November 2005; accepted 29 November 2005
Available online 3 April 2006

Abstract

In this study treatment of palm oil mill effluent (POME) was investigated using aerobic oxidation based on an activated sludge process.
The effects of sludge volume index, scum index and mixed liquor suspended solids during the acclimatizing phase and biomass build-up
phase were investigated in order to ascertain the reactor stability. The efficiency of the activated sludge process was evaluated by treating
anaerobically digested and diluted raw POME obtained from Golden Hope Plantations, Malaysia. The treatment of POME was carried
out at a fixed biomass concentration of 39007200 mg/L, whereas the corresponding sludge volume index was found to be around
10575 mL/g. The initial studies on the efficiency of the activated sludge reactor were carried out using diluted raw POME for varying the
hydraulic retention time, viz: 18, 24, 30 and 36 h and influent COD concentration, viz: 1000, 2000, 3000, 4000 and 5000 mg/L,
respectively. The results showed that at the end of 36 h of hydraulic retention time for the above said influent COD, the COD removal
efficiencies were found to be 83%, 72%, 64%, 54% and 42% whereas at 24 h hydraulic retention time they were 57%, 45%, 38%, 30%
and 27%, respectively. The effectiveness of aerobic oxidation was also compared between anaerobically digested and diluted raw POME
having corresponding CODs of 3908 and 3925 mg/L, for varying hydraulic retention time, viz: 18, 24, 30, 36, 42, 48, 54 and 60 h. The
dissolved oxygen concentration and pH in the activated sludge reactor were found to be 1.8–2.2 mg/L and 7–8.5, respectively. The scum
index was found to rise from 0.5% to 1.9% during the acclimatizing phase and biomass build-up phase.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Aerobic process; Activated sludge process; Suspended growth; Palm oil mill effluent; Scum index; Sludge volume index

1. Introduction rich in organic carbon with a biochemical oxygen demand

(BOD) higher than 20 g/L and nitrogen content around
The palm oil industry is one of the major agro-industries 0.2 g/L as ammonia nitrogen and 0.5 g/L total nitrogen (Ma
in Malaysia. The commercial palm oil production is mainly et al., 2001). The various effluent treatment schemes which
located in Peninsular Malaysia and dates back to the 1960s. are currently used by the Malaysian palm oil industry are
During the period between 1990 and 2002 palm oil listed in descending order: (a) anaerobic/facultative ponds
production was nearly doubled from 6,094,622 to (Rahim and Raj, 1982; Wong, 1980; Chan and Chooi, 1982),
11,880,000 (MT) per year, making Malaysia the biggest (b) tank digestion and mechanical aeration, (c) tank digestion
palm oil producer worldwide (http://www.malaysiainfor- and facultative ponds, (d) decanter and facultative ponds, and
mation.com/industries/palmoilindustry.htm). (e) physico-chemical and biological treatment (Andreasen,
The palm oil mill effluent (POME) is generated from three 1982). Treatment of POME has also been tried using
major sources, namely sterilizer condensate, hydrocyclone evaporation technology (Ma, 1999a) and a clarification
waste and separator sludge (Borja and Banks, 1994). On an system coupled with filtration and aeration (UNEP, 1994).
average 0.9–1.5 m3 of POME is generated for each ton of Pretreatment of POME was carried out in order to remove
crude palm oil produced (Davis and Reilly, 1980). POME is the suspended solids and residual oil using flocculation,
solvent extraction, adsorption and membrane separation
Corresponding author. Tel.: +603 8946 6416; fax: +603 8946 6425. process (Abdul Latif et al., 2003a). Adopting coagulation and
E-mail address: [email protected] (K. Vijayaraghavan). activated carbon adsorption as a pretreatment stage for

0301-4797/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jenvman.2005.11.016
 ARTICLE IN PRESS
K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31 25

POME treatment resulted in removal efficiencies of 97.9%

150 mm
5
turbidity, 56% COD and 71% BOD. The pretreated effluent, 6
2 3
when subjected to a membrane separation technique based on
ultra-filtration and reverse osmosis, resulted in the following
removal efficiencies: turbidity 100%, COD 98.8% and BOD

350 mm
1
99.4% (Abdul Latif et al., 2003b). Today 85% of POME
4
treatment is based on an anaerobic and facultative ponding
system, which is followed by another system consisting of an 400 mm
open tank digester coupled with extended aeration in a pond 650 mm
(Ma, 1999b). Due to global warming, these open types of
1. Feed tank 4. Diffused aeration
digesters are currently being converted into closed digesters. 2. Feed pump 5. Activated sludge reactor
Since methane is a green house gas, its release into the 3. Inlet 6. Outlet
environment is curtailed and the fuel value of the methane is Fig. 1. Experimental set-up of activated sludge reactor.
extracted by using it as an auxiliary fuel in the boilers of palm
oil mill. Chin and Wong (1981) and Ludzack (1960), treated
POME using a physico-chemical process followed by an
activated sludge process. rectangular shape with the following dimensions: 650 mm
Biological treatment of green olive debittering wastewater length  400 mm width  350 mm liquid depth. The aera-
was investigated using aerobic and anaerobic systems based tion was carried out using a diffused aeration system.
on combined and separate reactors. The results showed that
aerobic treatment was found to be more efficient than the
2.2. Sample preservation
anaerobic, whereas combined anaerobic–aerobic treatment
yielded less aerobic sludge (Aggelis et al., 2001). Treatment of
The POME was preserved at a temperature less than
vegetable oil refining industry wastewater based on coagula-
4 1C, but above the freezing point in order to prevent the
tion and a biological method resulted in an overall removal
wastewater from undergoing biodegradation due to micro-
efficiency of 92–96% COD, 83–98% TSS and 93–95% oil
bial action (APHA, 1985).
and grease, respectively (Azbar and Yonar, 2004).
Sludge bulking in activated sludge processes results due to
nutrient deficiency, especially nitrogen. A BOD/N ratio of 2.3. Palm oil mill effluent characteristics
100:4 resulted in better sludge settling, while ratios less than
100:3 resulted in poor settling due to the growth of filamentous The raw POME was collected from the collection pit of
organisms (Peng et al., 2003). The sludge bulking can be Golden Hope Plantation, Banting, Malaysia; its character-
controlled by subjecting it to intermittent aeration and main- istic is presented in Table 1. The effluent treatment scheme
taining the F/M ratio at 1.8–2.7 kg.BOD5/kg MLVSS d
1and in Golden Hope Plantation consisted of a collection pit
80–120 mL/g SVI for food waste (George and Andrew, 2003). followed by a closed-type anaerobic digester and an
The oxygen transfer rate in an activated sludge process also aerobic pond. The average effluent characteristics from
depends on the solids content. The fine bubble method of the anaerobic digester to the aerobic pond are presented in
diffused aeration was found to be economically effective with Table 2. The efficiency of the activated sludge reactor in
respect to power density and specific air throughput for an treating POME was investigated by using samples from the
mixed liquor suspended solids (MLSS) concentration up to anaerobic digester and diluted raw POME. The dilution of
18,000 mg/L (Krampe and Krauth, 2003). the raw POME was done in order to match the COD with
In the present scenario of POME treatment, anaerobic that of the effluent coming out of the anaerobic digester.
digestion was followed by aerobic oxidation in ponds.
Hence, in this article, the activated sludge process is
proposed as a post-treatment after the anaerobic diges- Table 1
tion process in order to curtail the large land area required Characteristics of raw palm oil mill effluent
by the aerobic pond. The efficiency of the activated sludge Parameters Concentrationa Standard deviation
process was evaluated by treating anaerobically digested
and diluted raw POME obtained from the effluent PH 3.570.1
treatment plant of Golden Hope Plantations, Malaysia. BOD5 25,545 70.1
COD 55,775 88.9
Suspended solids 18,479 102.2
2. Material and methods Total nitrogen 711 23.7
Ammoniacal nitrogen 36 5.0
2.1. Reactor set-up Oil and grease 8020 17.4
Temperature 8471
The schematic diagram of the activated sludge reactor is a
Except pH and temperature all other parameters are in mg/L,
shown Fig 1. The activated sludge reactor had a temperature in 1C.
 ARTICLE IN PRESS
26 K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31

Table 2 Table 3
Characteristics of anaerobically digested palm oil mill effluent Composition of nutrient solution

Parameters Concentrationa Standard deviation Parameters Concentration (mg/L)

pH 7.870.2 Ni 0.05
BOD5 25,545 70.1 Co 0.03
COD 1372 88.9 K 1.0
Suspended solids 512 102.2 Fe 1.0
Total nitrogen 134 23.7 Ca 2.0
Ammoniacal nitrogen 36 5.0
Oil and grease 8020 17.4
Temperature 2871 week, till it reached the full operating liquid level of about
a
Except pH and temperature all other parameters are in mg/L, 91 L capacity.
temperature in 1C. The acclimatizing operation was conducted under batch
operating conditions. During the acclimatization phase the
In order to vary the influent COD concentration to the activated sludge reactor was monitored with respect to pH,
activated sludge reactor, a suitable aliquot of raw palm MLSS, SI and SVI. The influent pH to the activated sludge
oil effluent was diluted and analyzed for its initial reactor was maintained in the neutral pH range (770.2).
characteristics.
2.5.3. Biomass build-up phase
2.4. Analytical methods At the end of the acclimatization phase the biomass
concentration (MLSS) was found to be 2370 mg/L, which
The organic strength (COD) of the wastewater was was below the desired biomass concentration for operating
determined by a colorimetric method (HACH, 1997). The the activated sludge process. Hence the biomass build-up
biodegradability of the wastewater was measured in terms phase was carried out in order to maintain a consistent
of BOD5. The biomass settling characteristics were MLSS concentration of about 39007200 mg/L. The
determined based on the sludge volume (Imhoff cone @ biomass concentration in the activated sludge reactor was
30 min settling period) and sludge volume index (SVI). The increased by operating the reactor under continuous feed
MLSS and MLVSS was determined by drying the sample mode with POME at 5000 mg/L COD at an HRT of 1.5 d.
at 105 1C and 550750 1C. The oil & grease and total The dosing of nutrients (Table 3) was carried out during
nitrogen were determined by the Soxhlet and Kjeldhal the first 2 days of the biomass build-up phase, thereafter a
methods (APHA, 1985). The scum index (SI) was used to second and final dosage was given after a period of 30 days.
ascertain microbial foaming (Torpak, 2006). During this phase the pH, SVI, MLSS and mixed liquor
volatile suspended solids (MLVSS) concentration in the
2.5. Start-up of activated sludge process reactor were monitored. Thereafter the treatment of
POME was investigated in the activated sludge reactor
The start-up of the activated sludge reactor was carried using anaerobically digested and diluted raw POME.
out in a three-phased manner consisting of a seeding phase,
an acclimatization phase and a biomass build-up phase. 3. Results and discussion

2.5.1. Seeding phase 3.1. Start-up

During this phase the initial seeding of microflora was
carried out using a mixture of activated sludge (from the The SVI and MLSS concentration during the acclimatiz-
sewage treatment plant) and cow dung at a 1:1 ratio in ing phase and biomass build-up phase while treating
volume. An initial volume of 20 L of seed culture POME are shown in Fig 2. During the acclimatizing phase
(microflora) with a solids content of 2.970.1% was fed the SVI and MLSS showed a steady rise and reached values
into the reactor. The reactor was supplemented with of 50 mL/g and 2370 mg/L, respectively. As shown in Fig 2
nutrients like Ni, Co, K, Fe and Ca whose concentrations the acclimatization phase was completed on the 8th week;
are presented in Table 3. during this period the MLSS concentration was found to
be 2370 mg/L which was relatively low when compared to
2.5.2. Acclimatization phase the desired MLSS concentration for operating the acti-
The seed microflora were acclimatized with POME by vated sludge reactor. New York Water Operators and
adding 10 L of wastewater consisting of POME and cow Metcalf & Eddy Inc (2003) stated a desired biomass
dung in the ratio (9:1) for a hydraulic retention time of 1 concentration of 3000–6000 mg/L MLSS and SVI of
week. The POME had an inlet organic strength of 3000 mg/ 50–180 mL/g for effective functioning of an activated
L COD while the cow dung had an inlet solids content of sludge reactor. Jenne et al. (2003) stated that bulking
2.970.1%. The addition of POME and cow dung was resulted when the SVI exceeded 150 mL/g. In order to
repeated in the ratio 9:1 by volume (9+1 L) at an HRT of 1 build-up the biomass concentration in the activated sludge
 ARTICLE IN PRESS
K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31 27

400 5000 5000 3

SVI MLSS
Biomass Build-up Phase Biomass Build-up Phase
MLSS Scum Index
350
2.5
4000 4000
300 Acclimatizing Phase
Acclimatizing Phase
2

Scum Index (%)

MLSS (mg/L)
250

MLSS (mg/L)
3000 3000
SVI (mL/g)

200 1.5

2000 2000
150
1

100
1000 1000
0.5
50

0 0
0 0
5 6 7 8 9 10 11 12 13 14 15 16 17 18 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Start-up period in Weeks Start-up period in Weeks

Fig. 2. Sludge volume index and MLSS concentration during the start-up Fig. 3. MLSS concentration and scum index during the start-up period of
period of activated sludge reactor. activated sludge reactor.

reactor, 60% of the clarified sludge was recycled back into to the desired COD concentration. The investigation was
the reactor. As the biomass recirculation was carried out carried out at varying influent COD values, namely 1000,
once every 2 h, a rise in SVI and MLSS concentration was 2000, 3000, 4000 and 5000 mg/L. The reason for fixing the
noticed in the activated sludge reactor. At the end of the higher limit of COD at 5000 mg/L to the activated sludge
16th week, the SVI and MLSS values were found to be reactor is that the effluent discharged from the anaerobic
100 mL/g and 3960 mg/L, respectively. The reason for digester of the Golden Hope Plantation into the aerobic
maintaining the MLSS concentration through biomass pond had an organic strength of 3908 mg/L COD.
recirculation is to have a constant biomass concentration The effect of COD removal versus hydraulic retention
(bacterial population) for the biological reaction to take time is shown in Fig 4. The results showed that as the HRT
place during the aerobic degradation of the POME. increased the corresponding COD removal efficiency also
Moreover, the biomass exhibited a good settling character- increased. The investigation was carried out under varying
istic as the SVI was found to be 100 mL/g. hydraulic retention times viz 18, 24, 30 and 36 h for varying
The relationship between MLSS and SI during the influent COD concentrations. At the end of 36 h of HRT
acclimatization and biomass build-up is shown in Fig 3. for influent CODs of 1000, 2000, 3000, 4000 and 5000 mg/L
The biomass concentration increased steadily from the the COD removal efficiencies were 83%, 72%, 64%,
acclimatization phase (5th week 1700 mg/L) to the biomass 54%and 42%, whereas at 24 h HRT the COD removal
build up phase (18th week 4150 mg/L). During the efficiencies were 57%, 45%, 38%, 30% and 27%,
acclimatization and biomass build up phase the SI also respectively. The removal of COD is reflected in the rise
showed a rise in value from 0.5% to 1.9%. The level of in biomass concentration, as the dissolved organics were
biological foaming was determined based on the SI. The converted into new cells.
maximum value of SI was found to be 1.9% during the The effect of BOD5 removal versus hydraulic retention
biomass build-up phase while treating POME in the time is shown in Fig 5. The results showed that at the end
activated sludge reactor. Torpak (2006) stated that a SI of 36 h HRT for influent BOD5 values of 440, 895, 1300,
in the range 0.5–6 resulted in less foaming problems. 1800 and 2300 mg/L the BOD5 removal efficiencies were
Though the filamentous microorganisms are believed to 74%, 66%, 61%, 42% and 39%, whereas at 24 h HRT the
form the ‘‘backbones’’ of activated sludge flocs on which BOD5 removal efficiencies were 54%, 41%, 35%, 28% and
floc forming bacteria are attached due to the secretion of 26%, respectively.
extracellular polymers, the overgrowth of the filamentous The effect of oil and grease removal versus hydraulic
microorganisms will lead to foaming and sludge bulking retention time is shown in Fig 6. For influent oil and grease
problems in the activated sludge process. As the present concentrations of 144, 287,431, 575 and 720 mg/L at an
investigation was free from sludge bulking and foaming, it HRT of 36 h the oil and grease removal percentages were
can be concluded that the filamentous microorganisms found to be 53%, 45%, 39%, 34% and 32%, whereas at
were low in number. 24 h HRT the removal efficiencies were found to be 33%,
26%, 22%, 20% and 18%, respectively. The possible
3.2. Treatment of diluted raw palm oil mill effluent in the reason for the decrease in oil content could be due to the
activated sludge reactor aerobic degradation of the oil and to some extent
adsorption on the biomass.
The treatment of POME in activated sludge was carried The residual total Kjeldhal nitrogen versus hydraulic
out by taking suitable aliquots of raw POME and diluting retention time is shown Fig. 7. The results showed that at an
 ARTICLE IN PRESS
28 K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31

100 45
Initial COD Concentration (mg/L) Initial TKN Concentration (mg/L)
40 12
1000
24
2000
35 39
75 3000
53
4000

Residual TKN (mg/L)

30 65
5000
COD removal (%)

25
50
20

15

25 10

0
0 10 15 20 25 30 35 40
10 15 20 25 30 35 40
Hydraulic Retention Time (hr)
Hydraulic Retention Time (hr)
Fig. 7. Residual TKN versus hydraulic retention time during aerobic
Fig. 4. COD removal percent versus hydraulic retention time during oxidation of diluted raw palm oil mill effluent.
aerobic oxidation of diluted raw palm oil mill effluent.

100
HRT of 36 h for an influent TKN of 12, 24, 39, 53 and
Initial BOD Concentration (mg/L) 63 mg/L the residual TKN concentrations were found to be
440 0.7, 1.5, 7, 12 and 15 mg/L, whereas at 24 h HRT the
895
1300 residual TKN concentrations were 7, 8, 17, 21 and 25 mg/L,
75
1800 respectively. The possible reason for the decrease in
2300
BOD removal (%)

nitrogen concentration in the treated POME could be due

to the utilization of nitrogen in cell growth by the biomass.
50
The dissolved oxygen (DO) concentrations in the
activated sludge reactor were found to be between 1.8
and 2.2 mg/L. (US EPA, 1993) states that as the dissolved
25 oxygen concentration falls below 2–3 mg/L the activity of
nitrifying bacteria will be reduced or totally inhibited. Since
the activated sludge reactor biomass was free from bulking
0 and foaming and had a DO concentration between 1.8 and
10 15 20 25 30 35 40 2.2 mg/L, it could be presumed that the nitrifying bacterial
Hydraulic Retention Time (hr)
population was relatively low.
Fig. 5. BOD removal percent versus hydraulic retention time during
aerobic oxidation of diluted raw palm oil mill effluent. 3.3. Comparison of treatment efficiency between
anaerobically digested and diluted palm oil mill effluent in
activated sludge reactor
60
Initial Oil & Grease Concentration (mg/L)
The efficiency of the activated sludge reactor was
144
50 287 measured in terms of organic removal by feeding anaero-
431 bically digested POME and diluted raw POME having a
Oil & Grease removal (%)

575
40 720
COD similar to that of anaerobically digested effluent.
Figs. 8 and 9 show the comparison between COD and
BOD5 removal efficiencies when the influent to the
30
activated sludge reactor was anaerobically digested and
diluted raw POME. The COD and BOD5 removal
20
efficiencies increased with the HRT irrespective of the
source of influent, but the anaerobically digested sample
10 resulted in higher COD and BOD5 removal efficiencies
than the diluted raw POME. The influent concentration of
0 the anaerobically digested POME was found to be
10 15 20 25 30 35 40 3908 mg/L, while the raw POME was diluted to the range
Hydraulic Retention Time (hr)
of the anaerobic digested sample resulting in a COD of
Fig. 6. Oil and grease removal percent versus hydraulic retention time 3925 mg/L. The aerobic oxidation process for the above
during aerobic oxidation of diluted raw palm oil mill effluent. said influent COD concentrations at 60 h HRT resulted in
 ARTICLE IN PRESS
K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31 29

120 500

Anaerobically digested POME Anaerobically digested POME

450
Initial COD 3908 mg/L Initial Oil & Grease: 237 mg/L
100
400 Diluted raw POME

Residual Oil & Grease (mg/L)

Diluted raw POME
Initial Oil & Grease: 558 mg/L
Initial COD 3925 mg/L 350
80
COD removal (%)

300

60 250

200
40
150

100
20
50

0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70
Hydraulic Retention Time (hr) Hydraulic Retention Time (hr)

Fig. 8. COD removal percent versus hydraulic retention time during Fig. 10. Residual oil and grease versus hydraulic retention time during
aerobic oxidation of anaerobically digested and diluted raw palm oil mill aerobic oxidation of anaerobically digested and diluted raw palm oil mill
effluent. effluent.

100 concentration in the case of anaerobically digested POME

90 Anaerobically digested POME could be due to the degradation under anaerobic condi-
Initial BOD: 1720 mg/L
80 tions and a fraction being accumulated within the
Diluted raw POME
Initial BOD: 1783 mg/L anaerobic digester as a scum. In the case of diluted raw
70
POME a more homogenous composition resulted with
BOD removal (%)

60
respect to the original value, as it had not been subjected to
50 any form of treatment. Hence the concentration of oil and
40 grease was found to be higher in the diluted POME in
30
comparison to the anaerobically digested POME. At the
end of 60 h of aerobic oxidation the respective residual oil
20
and grease concentrations were found to be 24 and 112 mg/
10 L, respectively. The results clearly showed that during
0 anaerobic digestion, cleavage in the long chains of the oil
0 10 20 30 40 50 60 70
Hydraulic Retention Time (hr)
molecules could have occurred leading to easy degradation
in the subsequent aerobic stage, leading to low residual
Fig. 9. BOD removal percent versus hydraulic retention time during concentration in the treated effluent of the activated sludge
aerobic oxidation of anaerobically digested and diluted raw palm oil mill
reactor.
effluent.
Fig 11 shows the residual Kjeldhal nitrogen concentra-
tion versus hydraulic retention time. The influent TKN
concentration to the activated sludge reactor varied
98% and 89% of COD removal, respectively. In the case of depending on the sample. In the case of the anaerobically
BOD5 the influent concentrations for the anaerobically digested sample the influent TKN concentration was found
digested sample and the diluted sample were found to be to 319 mg/L, whereas it was 53 mg/L for diluted raw
1720 and 1753 mg/L. At the end of 60 h of aerobic POME. The higher TKN concentration in the anaerobic
oxidation the BOD5 removal percentages were found to sample when compared to the diluted raw POME could be
be 93% and 82%, respectively. The possible reason for the due to the following reason. In the case of the anaerobically
higher organic removal in the anaerobically digested digested sample there was no dilution, so the removal of
sample could be due to the presence of partially degraded nitrogen from the effluent could have occurred due to cell
organics making then more amenable to aerobic digestion. synthesis and loss as ammonia under anaerobic conditions.
In the case of the diluted raw palm oil effluent, the sample In the case of diluted raw POME a lower TKN
was not subjected to prior degradation. concentration was observed, as the dilution of raw POME
Fig. 10 shows the residual oil and grease concentration was carried out in order to meet the COD value of
versus hydraulic retention time during the aerobic oxida- anaerobically digested effluent.
tion of anaerobically digested and diluted raw POME as As shown in Fig 11, the nitrogen content of the
feed. In the case of the anaerobically digested sample the anaerobically digested and diluted raw POME decreased
inlet oil and grease concentration was found to be 237 mg/ from an initial value of 319–58 mg/L and from an initial
L, while the diluted raw POME had an inlet concentration value of 53–3 mg/L, respectively, at the end of 60 h of
of 558 mg/L. The reason for the low inlet oil and grease hydraulic retention time. The reduction in the nitrogen
 ARTICLE IN PRESS
30 K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31

350
48, 54 and 60 h, respectively. Aerobic oxidation of
Anaerobically digested POME
Initial TKN: 319 mg/L
anaerobically digested POME, at a hydraulic retention
300
Diluted raw POME time of 60 h, resulted in COD removal efficiencies of 98%
250
Initial TKN: 53 mg/L and 93% with residual oil and grease and TKN values of
Residual TKN (mg/L)

24 and 58 mg/L, respectively. In the case of aerobic

200 oxidation of diluted raw POME at the end of 60 h HRT,
the COD and BOD5 removals were found to be 89% and
150 82% with residual oil and grease and TKN values of 112
and 3 mg/L, respectively. The dissolved oxygen concentra-
100
tion, pH and SI ranged between 1.8 and 2.2 mg/L, 7% and
8.5% and 1.1% and 1.9%, respectively, during the
50
operation of the activated sludge reactor.
0
0 10 20 30 40 50 60 70
Hydraulic Retention Time (hr)
References
Fig. 11. Residual TKN versus hydraulic retention time during aerobic
oxidation of anaerobically digested and diluted raw palm oil mill effluent. Abdul Latif, A., Suzylawati, I., Norliza, I., Subhash, B., 2003a. Removal
of suspended solids and residual oil from palm oil mill effluent. Journal
of Chemical Technology and Biotechnology 78 (9), 971–978.
Abdul Latif, A., Suzylawati, I., Norliza, I., Subhash, B., 2003b. Water
level could be due to the synthesis of new cells and a recycling from palm oil mill effluent using membrane technology.
fraction of the nitrogen could have been lost to the Desalination 157, 87–95.
environment as ammonia while maintaining the equili- Aggelis, G.G., Gavala, H.N., Lyberatos, G., 2001. Combined and
brium [NH3
N ¼ NH3+NH+ 4 ] (Eugene, 2000). The
separate aerobic and anaerobic bio-treatment of green olive debittering
activated sludge reactor was operated in a pH range of wastewater. Journal of Agricultural Engineering Research 80 (3),
283–292.
7–8.5. The observed operating pH was in agreement with Andreasen, T., 1982. The AMINODAN system for treatment of
those of other researchers. For example, Xianling et al. palm oil mill effluent. In: Proceedings of Regional Workshop on
(2005) reported a pH value of 7.0–8.0 during aerobic Palm Oil Mill Technology and Effluent Treatment. PORIM, Malaysia,
treatment of refinery wastewater using an airlift loop pp. 213–215.
reactor. Venkata Mohan et al. (2005) reported a pH value APHA, 1985. Standard Methods for the Examination of Water and
Wastewater. 16th ed. Washington, DC.
of 7.5–7.9 for effective functioning of a sequencing batch Azbar, N., Yonar, T., 2004. Comparative evaluation of a laboratory and
reactor. full-scale treatment alternatives for the vegetable oil refining industry
wastewater. Process Biochemistry 39, 869–875.
4. Conclusion Borja, R., Banks, C.J., 1994. Anaerobic digestion of palm oil mill effluent
using up-flow anaerobic sludge blanket reactor. Biomass & Bioenergy
6, 381–389.
Aerobic oxidation of anaerobically digested and diluted Chan, K.S., Chooi, C.F., 1982. Ponding system for palm oil mill effluent
raw POME was investigated using an activated sludge treatment. In: Proceedings of Regional Workshop on Palm Oil Mill
reactor. Treatment of POME was carried out at a biomass Technology and Effluent Treatment. PORIM, Malaysia, pp. 185–192.
concentration of 39007200 mg/L, whereas the correspond- Chin, K.K., Wong, K.K., 1981. Palm oil refinery waste treatment. Water
ing SVI was found to be around 10575 mL/g. The initial Research 15, 1087–1092.
Davis, J.B., Reilly, P.J.A., 1980. Palm oil mill effluent—a summary of
studies on the efficiency of the activated sludge reactor treatment methods. Oleagineux 35, 323–330.
were carried out using diluted raw POME at varying Eugene, R.W., 2000. Applications of Environmental Chemistry—A
hydraulic retention times, viz: 18, 24, 30 and 36 h and Practical Guide for Environmental Professionals. CRC Press, USA.
influent COD concentrations, viz: 1000, 2000, 3000, 4000 George, N., Andrew, L., 2003. Control of filamentous organisms in food
processing wastewater treatment by intermittent aeration and selec-
and 5000 mg/L, respectively. The results showed that at the
tors. Journal of Chemical Technology & Biotechnology 78 (4),
end of 36 h of hydraulic retention time for the above-said 420–430.
influent COD concentrations, the COD removal efficien- HACH, 1997. Wastewater analysis user manual, In house edition.
cies were found to be 83%, 72%, 64%, 54% and42% http://www.malaysiainformation.com/industries/palmoilindustry.htm.
whereas at 24 h hydraulic retention time they were 57%, (accessed on 2004).
45%, 38%, 30% and 27%, respectively. At a HRT of 36 h, Jenne, R., Banadda, E., Philips, N., Van Impe, J.F., 2003. Image analysis
as a monitoring tool for activated sludge properties in lab-scale
for influent TKN concentrations of 12, 24, 39, 53 and installations. Environmental Science and Health Part-A 38 (10),
65 mg/L the residual TKN concentrations were 0.7, 1.5, 7, 2009–2018.
12 and 15 mg/L, whereas at 24 h HRT the residual TKN Krampe, J., Krauth, K., 2003. Oxygen transfer into activated sludge with
concentrations were 7, 8, 17, 21 and 25 mg/L, respectively. high MLSS concentrations. Water Science and Technology 47 (11),
The effectiveness of aerobic oxidation was also compared 297–303.
Ludzack, F.J., 1960. Laboratory model activated sludge unit. Journal of
between anaerobically digested and diluted raw POME Water Pollution Control Federation 605.
having corresponding COD values of 3908 and 3925 mg/L, Ma, A.N., 1999a. Innovations in management of palm oil mill effluent.
for varying hydraulic retention times, viz: 18, 24, 30, 36,42, The Planter, Kuala Lumpur 75 (881), 381–389.
 ARTICLE IN PRESS
K. Vijayaraghavan et al. / Journal of Environmental Management 82 (2007) 24–31 31

Ma, A.N., 1999b. Treatment of palm oil mill effluent. In: Gurmit, Singh, Torpak, H., 2006. Activated sludge http://web.deu.edu.tr/atiksu/toprak/
et al. (Eds.), Oil Palm and the Environment. A Malaysian Perspective. ani406.html (accessed on 2006).
MOPGC, Kuala Lumpur, pp. 113–125. UNEP, 1994. Treatment wastewater in the rubber industry. In: Cleaner
Ma, A.N., Chow, C.S., John, C.K., Ibrahim, A., Isa, Z., 2001. Palm oil Production in the Asia-Pacific Economic Cooperation Region. United
mill effluent—a survey. In: Proceedings of the PORIM Regional Nations Environment Programme, Paris, pp. 21–30.
Workshop on Palm Oil Mill Technology and Effluent Treatment. US EPA, 1993. Process Design Manual for Nitrogen Control, EPA/625/
Palm Oil Research Institute of Malaysia (PORIM) Serdang, Malaysia, R-93/010. Washington, DC.
pp. 233–269. Venkata Mohan, S., Chandrashekara Rao, N., Krishna Prasad, K.,
Metcalf & Eddy Inc, 2003. Wastewater engineering treatment & reuse. In: Madhavi, B.T.V., Sharma, P.N., 2005. Treatment of complex chemical
Tchobanoglous, G., Burton, F.L., Stensel, H.D. (Eds.), Tata McGraw- wastewater in a sequencing batch reactor (SBR) with an aerobic
Hill Publishing Company Limited, fourth ed., New Delhi, India. suspended growth configuration. Process Biochemistry 40 (5),
New York Water Operators. http://water.me.vccs.edu/ (accessed on 2005). 1501–1508.
Peng, Y., Gao, C., Wang, S., Ozaki, M., Takigawa, A., 2003. Non- Wong, F.M., 1980. A review on the progress of compliance with the palm
filamentous sludge bulking caused by a deficiency of nitrogen in industrial oil control regulations. Seminar on Advances in Palm Oil Effluent
wastewater treatment. Water Science and Technology 47 (11), 289–295. Control Technology, Kuala Lumpur, pp. 142–149.
Rahim, B.A., Raj, R., 1982. Pilot plant study of a biological treatment Xianling, L., Jianping, W., Qing, Y., Xueming, Z., 2005. The pilot study
system for palm oil mill effluent. In: Proceedings of Regional for oil refinery wastewater treatment using a gas–liquid–solid three-
Workshop on Palm Oil Mill Technology and Effluent Treatment. phase flow airlift loop bioreactor. Journal Biochemical Engineering 27,
PORIM, Malaysia, pp. 163–170. 40–44.