Phenol removal from oil refinery wastewater using anaerobic stabilization pond modeling and process optimization using response surface methodology (RSM)

Full text

*Corresponding author.
1944-3994 / 1944-3986 © 2017 Desalination Publications. All rights reserved.
Desalination and Water Treatment
www.deswater.com
doi:10.5004/dwt.2017.21064
87 (2017) 199–208
August
Phenol removal from oil refinery wastewater using anaerobic stabilization pond
modeling and process optimization using response surface methodology (RSM)
Abdollah Dargahia, Mitra Mohammadib, Farhad Amirianc, Amir Karamib, Ali Almasib,*
aDepartment of Environmental Health Engineering, School of Public Health, Hamadan University of Medical Sciences, Hamadan, Iran,
Tel. +989141597607, email: a.dargahi29@yahoo.com (A. Dargahi)
bDepartment of Environmental Health Engineering, School of Public Health, Kermanshah University of Medical Sciences,
Kermanshah, Iran, Tel. +989188573064, email: m.mohamadi725@ gmail.com (M. Mohammadi), Tel. +989183598931,
emal: amir_karami119@yahoo.com (A. Karami), Tel. +989181317314, email: alialmasi@yahoo.com (A. Almasi)
cDepartment of Pathology, School of Medicine, Kermanshah University of Medical Sciences, Kermanshah, Iran, Tel. +989339755662,
email: farhad_amirian @yahoo.com (F. Amirian)
Received 10 October 2016; Accepted 24 June 2017
a b s t r a c t
Oil refinery process releases toxic pollutants into aqueous environment. Phenol and its derivations as
the most important pollutants pose severe environmental concern. In this study, the rectangle anaer-
obic stabilization pond (ASP) consisting of feed tank with workload of 60 Lit (1 × 0.2 × 1) meter of
phenol was used. This study evaluated the interactive effect of phenol concentration (200–400 mg/l),
temperature (8–24°C) and Hydraulic retention time of (HRT) (2–5 d) on the efficiency of anaerobic
stabilization pond for oil refinery wastewater treatment. In this study, experiments were carried out
based on central composite design (CCD) and analyzed and modeled by response surface method-
ology (RSM) aimed at demonstrating the operating variables and also the interactive effect of three
independent variables on 7 responses. The maximum removal efficiency of SCOD, TCOD, SBOD and
TBOD were 66.26, 68.95, 65.3 and 67.02%, respectively, at phenol concentration of 200 mg/L, HRT of
2 d, and temperature of 24°C. Generally, the ratio of N/P varied between 6.69–9.12 and 7.04–12.93,
respectively, in influent and effluent of anaerobic stabilization pond. The maximum phenol removal
efficiency reached 70.53% and 81.63% at phenol concentration of 200 mg/L, temperature of 24°C with
HRT (2 and 5 d), respectively. The phenol removal efficiency was significantly influenced by increas-
ing the temperature compared to decreasing the phenol concentration. The result indicated that the
anaerobic stabilization pond was a capable biological treatment process that could achieve the mod-
erate removal of oil refinery wastewater.
Keywords: Phenol; Oil refinery; Wastewater; Anaerobic stabilization pond; RSM
1. Int roduction
Crude oil refinery plants are considered among the envi-
ronmental polluting industries that discharge annually large
volume of effluent from various processing units into waste-
water treatment plants [1]. In Petroleum refinery, crude oil
is processed and transformed into 2500 more useful refined
products including, diesel fuel, gas oil, gasoline, kerosene,
petroleum oils, etc [2]. The quantity and characteristics of
generated wastewater in oil refinery plant depends on the
type of processing unit that regularly contains the pollut-
ants such as phenol, benzene, heavy metal, nutrient , etc [3].
Industrial effluent contains toxic pollutants that are intro-
duced into aqueous environment and wastewater treatment
plant, repeatedly [4]. So far water pollution through organic
and inorganic substances as a result of industrial activities
in oil refinery has posed severe difficulties and challenges
for public health [5]. Among these pollutants, phenol is of
great importance, given that phenol and its derivations
have been considered as serious environmental concern that
causes extreme toxicity impact to human beings, air and

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
200
soil ecology as well [6]. Phenol inhibited the synthesis and
replication of DNA in cell, and additionally as noted in the
studies, it prevented from the reparation of DNA in diploid
human fibroblast [5]. Phenol has been placed in the list of
the most important pollutants according to Environmental
Protection Agency (EPA) and European Agency (EU) due to
its high toxicity, persistence, bio accumulative and improper
organoleptic characteristics [7]. Such a practical and environ-
mental-friendly technology should be considered for phe-
nol removal before discharging to the environment due to
actually numerous threats of phenol compounds to human
health and environment [8]. Currently, petrochemical refin-
ery industry emphasizes techniques particularly cost-effec-
tive and high-efficiency wastewater treatment technology
[9]. It should be taken into account that toxicity of residual
petroleum products, significant quantities of industrial efflu-
ent, poor biodegradability of oil compounds and leaching
base on environmental standard have created the major con-
cern [10]. The result of various studies revealed that micro-
organisms in biological systems are developed as bacterial
community against the environmental destructive effects,
like biomass against the wastewater toxicity [11].
Stabilization pond has been regarded as one of the
most efficient biological treatment processes, it is especially
appropriate in tropical and subtropical regions [12]. Gen-
erally, the efficiency of stabilization ponds depends on the
environmental condition such as ample sunlight and tem-
perature as the key factors and precipitation, evaporation
and presence of toxic pollutants as well [13]. Optimization
of the process variables is necessary to attain the optimal
removal efficiency. The usual experimental design needs
a large number of experiments resulting in some time and
economical problems. Therefore, it can be administered
using the statistical experiments design that minimizes the
number of experiments [14]. The objective was to optimize
the responses which were affected by independent vari-
ables and their interaction, given the experimental design.
In this study, the experiment was designed based on cen-
tral composite design (CCD) and was analyzed by response
surface methodology (RSM), that provided proper statisti-
cal tools aimed at designing and analyzing experiments for
process optimization [15].
Present study, modeled and analyzed oil refinery waste-
water treatment by anaerobic stabilization pond using RSM,
that analyses the simultaneous effect of three independent
variables (phenol concentration, HRT and temperature) on
7 parameters as responses (SCOD, TCOD, SBOD5, TBOD,
N-NH3, P-PO4 and Phenol.
2. Method and materials
2.1. Chemicals
All chemicals and reagents were prepared with purity
of 99.99% purchased from (Merck Co. Germany). It should
be noted that deionized water which was procured in lab-
oratory was used to prepare the desired standard solution.
2.2. Type of wastewater
Wastewater was procured from separator unit effluent
in oil refinery process plant, Kermanshah, Iran.
2.3. Start up and operation of anaerobic stabilization pond
A pilot scale of anaerobic stabilization pond was used
in this study, composed of feed tank with V workload of
60 L and equipped with valve disk aiming at flow adjust-
ment. The rectangle stabilization pond with dimensions of
1 × 0.2 × 1 m, was completely sealed to prevent from oxy-
gen entrance. The wastewater was fed into stabilization
pond with HRT (2 and 5 d), based on 95 and 40 L/d of
influent, respectively. The stabilization pond was seeded
with 1.5 L of sludge seed taken from municipal wastewa-
ter plant that had been equalized and mixed. The micro-
organism acclimation to achieve the steady state of system
was lasted for about 90 d. The stabilization pond was fed
in bath process by combination of molasses and refinery
oil wastewater aimed at regulating the organic load rate
of system. The molasses wastewater was provided by
sugar production factory, Kermanshah, Iran. The molas-
ses wastewater contains the COD and BOD5 levels of 2400
mg/L and 1680 mg/L, respectively.
2.4. Chemical analysis
In present study, composite sampling was done every
2 h, which was immediately transferred to laboratory and
was kept in refrigerator (2–4°C). Analytical laboratory test
was done on collected samples. Chemical oxygen demand
(COD) was measured based on closed reflex method (Jen-
way, Hach USA DR 5000), (Method 5220C). Biological oxy-
gen demand (BOD5) was measured for five days according
to titrimetric method (Method 5210B). pH was analyzed
using pH meter (Digimed model DM-20, Digicron Analíti-
caLtda, São Paulo, Brazil). N-NH4 was analyzed by direct
Nesslerization (Method 4500 C), Colorimetric method was
used to determine phosphate (Method 4500 P), and photo-
metric method was used for phenol analysis (Method 5530
C). Nitrate, phosphate and phenol were determined by
spectrophotometer (Varian UV-120-02 California, USA). UV
probe software was used to control the system and obtain
results of experiment. It should be noted that sample col-
lection, transfer and all chemical analysis were performed
according to standard method for water and wastewater
examination [16].
2.5. Experimental design
Central composite design, Box-Behnken design,
Hybrid design and three level factorial design are the
different classes of response surface design. RSM design
is the most frequently used Central composite design
(CCD). Aimed at analyzing the data, RSM was used as a
technique for collecting mathematical and statistical data
in order to evaluate three independent variables, that
is, phenol concentration (A), temperature (B) and HRT
(C), surrounded by phenol concentration of (200–400
mg/L), temperature (8–24°C), and HRT (2, 5 d) to assess
the 7 different responses; removal efficiency of TCOD
(total chemical oxygen demand), SCOD (soluble chem-
ical oxygen demand), TBOD5 (total biochemical oxygen
demand), SBOD5 (soluble biochemical oxygen demand),
N-NH3, P-PO4 and phenol). Accordingly, 13 experiments

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
201
were designed. The design consisted of (4 variable points,
4 axial points, 1 central point and 4 repetition points in
center). The result obtained based on CCD, was analyzed
using ANOVA variable analysis software, and multiple
regression analysis. Also the result can be demonstrated
by 3-D plot regarding the simultaneous effect of indepen-
dent variables on the responses. In this study, the rela-
tionships among response, input and quadratic equation
model were determined to predict the optimal variables
identified:
Y = β0+ βiXi+ βjXj+ βiiXi
2 + βjjXj
2 + βijXiXj + ... (1)
Y, i, j, b, X are process response, linear coefficient, quadratic
coefficient, regression coefficient and coded independent
variables, respectively.
3. Result and discussion
3.1. Statistical analysis
Central composite design (CCD) was selected to explore
the correlations between variables and responses. The
experimental data obtained from the 7 responses (Y1–Y7)
are demonstrated in (Tables 1 and 2). Based on experimen-
tal design (Tables 1 and 2), the response surface analysis
was fitted properly to the experimental result. Factor coded
models as well as analysis of variance (ANOVA) are rep-
resented for all responses in (Tables 3 and 4). The different
degree polynomial models are used for data fitting. Fitting
the experimental data was conducted by higher degree
polynomial equation i.e Quadratic vs 2FI and linear. The
final model terms were obtained after removing insignif-
icant variables and their interaction. The significance of
Table 1
Experimental conditions and results (HRT: 2 days)
Run Temp,
°C
Phenol.
Conc. mg/l
Rem.
SCOD, %
Rem.
TCOD, %
Rem.
SBOD5, %
Rem.
TBOD5, %
Rem.
N-NH3, %
Rem.
P-PO4, %
Rem.
Phenol, %
N/P
out
124 400 49.41 55.63 49.67 53.5 33.66 41.15 55.86 8.71
216 200 50.6 51.83 49.87 50.36 27.4 36.17 4 7. 9 6 12.5
324 300 57.14 62.8 59.83 61.69 39.03 46.41 63.47 8
416 300 41.3 45.1 43.22 42.69 23.1 29.15 41.19 8.75
516 300 39.2 4 7. 7 9 45.52 44.99 24.8 31.45 44.89 8.65
616 300 43.1 42.17 44.91 40.2 21.2 27.05 36.98 8.51
716 300 37.6 45 46.87 39.29 23 28.11 40.89 8.36
8 8 200 37. 7 1 38.72 35.18 36.94 15.92 17.92 31.39 12.16
9 8 300 29.14 31.41 28.6 29.7 13.13 15.9 24.39 9.12
10 16 300 41.8 4 7.1 41.87 44.98 19.9 30.02 42.98 8.37
11 24 200 66.26 68.95 65.3 67.02 43.39 58.23 70.53 12.93
12 8400 21.4 24.21 20.83 22.83 8.43 10.88 18.22 7.0 4
13 16 400 34.4 38.41 33.81 36.61 19.3 23.01 34.04 7. 68
Table 2
Experimental conditions and results (HRT: 5 days)
Run Temp,
°C
Phenol.
Conc. mg/l
Rem.
SCOD, %
Rem.
TCOD, %
Rem.
SBOD5, %
Rem.
TBOD5, %
Rem.
N-NH3, %
Rem.
P-PO4, %
Rem.
Phenol, %
N/P
out
124 400 58.67 62.12 56.96 59.13 40.21 49.73 61.12 8.71
216 200 54 57. 5 4 54.3 53.95 36.4 41.38 54.77 9.2
324 300 68.84 72.47 66.85 70.52 49.12 53.16 69.38 8.15
416 300 51.66 52.67 50.36 51.06 30.1 35.24 45.61 8.75
516 300 49.16 54.17 52.66 53.16 32.3 3 7. 34 44.31 8.53
616 300 54.16 51.1 49.6 3 48.86 27. 8 32.84 46.98 8.39
716 300 51.6 52.1 52.33 51 31.3 35.12 44.8 8.74
8 8 200 40.22 42.65 38.54 39.86 21.64 25.57 35.92 7.5
9 8 300 34.49 36.87 33.87 35.61 16.89 21.65 29.82 9.12
10 16 300 52 53.6 48.2 53.2 29.1 32.51 4 6.61 8.58
11 24. 200 73.51 76.44 71.84 73.32 56.22 65.19 81.63 10.8
12 8400 27. 3 6 26.63 26.61 28.38 12.66 16.74 21.24 7. 0 6
13 16 400 42.98 43.4 40.87 41.57 24.3 31.12 39.18 7.9

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
202
each model was identified using P value and F value. The
smaller amount of (P-values < 0.05) and great amount of
F-value show the greater significance of corresponding
model [19]. The statistical analysis showed the significance
of each model (P-value < 0.0001) for all responses at HRT of
(2, 5 days) that indicated the significance of the correspond-
ing model.
As it is shown in Tables 3 and 4, the F-values were
215.88, 189.6, 197.82, 207.66, 275.55, 303.83 and 174.93 at
HRT of 2 d, and 499.8, 121.29, 263.05, 196.28, 310.76, 588.75
and 256.08, respectively, at HRT of 5 d, for 7 responses (Y1–
Y7). Lack of fit (LOF) indicated the variation of data around
the fitted models, based on the result. In this study, the
LOF was not significant for all the models. It represents a
proper model predicted. Furthermore, the fitting of model
verified by Pred R-Squared, adjusted R2 and R2 between
the experimental and the model predicted values. In this
study, Pred R-Squared, Adjusted R2 and R2 are very close to
each other and nearly 1. The R2 in all models was remark-
ably more than (R2 > 0.98), in other words, about 0.98 of
variation for TBOD5, SBOD5, TCOD, SCOD, N-NH3, P-PO4
and phenol removal was explained by independent vari-
ables appropriately. The validity and reliability of anal-
ysis identified through the adequate values of 4 or more
[14]), were generally based on the result and the p-value
for all responses were significantly over 4 and between
44.58–56.73 and 34.4–75.94, respectively at HRT (2, 5 d). It
showed the proper value for analysis validated. Moreover,
the low amount of standard deviation at HRT (2, 5 d) were
3.32–5.83 and 1.84–6.51, respectively, and also coefficient
of variation (CV) were 1.4–1.98 and 0.96–2.37 at HRT (2,
5 d), respectively, representing the considerable reliability
and proper precision of the experiments [17]. It should be
mentioned that the coefficient and mathematical symbol
(+/-) of resulted equation showed the most effective vari-
able in anaerobic stabilization pond performance.
Table 3
ANOVA results for the equations of the Design Expert 6.0.6 for studied responses, HRT= 2 day
Response, % Modified equations in
terms of code factors
Type of
model
F value Adeq
precision
R2Adj.
R2
Pred.
R2
S.D. C .V PRESS Probability
for lack of fit
Rem. SCOD Y1 = +40.9 – 8.23A + 13.59
B + 2.4A2
Quadratic 174.93 46.02 0.983 0.977 0.974 1.71 4.07 44.59 0.8911
Rem. TCOD Y2 = +46.09 – 6.88A
+ 15.51 B
Linear 303.83 55.28 0.983 0.98 0.976 1.69 3.66 40.65 0.8992
Rem. SBOD5 Y3 = +44.39 – 7.67A +
15.03 B – 1.95A2
Linear 275.55 56.73 0.989 0.985 0.982 1.44 3.32 30.2 0.9684
Rem. TBOD5 Y4 = +42.40 – 6.90A +
15.46 + 2.88B2
Quadratic 207.6 6 48.14 0.985 0.981 0.978 1.67 3.83 38.32 0.9535
Rem. N-NH3Y5 = +22.67 – 4.22A +
13.10B + 2.92B2
Quadratic 19 7. 82 44.58 0.985 0.98 0.975 1.4 5.83 29.28 0.9508
Rem. P-PO4Y6 = +29.28 – 6.21A + 16.85
B + 2.47B2 – 2.51 AB
Quadratic 189.6 46.03 0.989 0.984 0.96 1.62 5.31 78.19 0.5785
Rem. Phenol Y7 = +41.28 – 6.96A +
19.31 B +2.70B2
Quadratic 215.88 4 7. 72 0.986 0.981 0.98 1.98 4.67 50.1 0.9996
Table 4
ANOVA results for the equations of the Design Expert 6.0.6 for studied responses, HRT = 5 day
Response, % Modified equations in
terms of code factors
Type of
model
F value Adeq
precision
R2Adj.
R2
Pred.
R2
S.D. C .V PRESS Probability
for lack of fit
Rem. SCOD Y1 = +50.67 – 6.45A +
16.49 B
Linear 256.08 49.83 0.98 0.977 0.969 1.92 3.78 58.82 0.4247
Rem. TCOD Y2 = +52.79 – 7.41A +
17.48B – 2.48A2 + 1.72B2
Quadratic 588.75 34.4 0.996 0.994 0.993 0.96 1.84 15.3 0.8905
Rem. SBOD5 Y3 = +50.56 – 6.71A +
16.10 B – 2.37A2
Quadratic 310.76 58.47 0.99 0.987 0.982 1.41 2.84 32.24 0.9370
Rem. TBOD5 Y4 = +51.92 – 6.34A +
16.52 B – 2.55A2
Quadratic 196.28 45.89 0.984 0.979 0.968 1.8 3.54 61.43 0.5223
Rem. N-NH3Y5 = +30.19 – 6.18A +
15.73 B + 2.60B2 – 1.76 AB
Quadratic 263.05 54.82 0.992 0.988 0.987 1.29 4.11 21.64 0.9915
Rem. P-PO4Y6 = +34.65 – 5.76A +
17.35 B + 4.02B2
Quadratic 121.29 35.03 0.975 0.967 0.942 2.38 6.51 122.07 0.4011
Rem. Phenol Y7 = +45.89 – 8.46A +
20.86 B + 3.96B2 – 1.46 AB
Quadratic 499.8 75..94 0.996 0.994 0.985 1.25 2 .61 45.93 0.2571

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
203
3.2. Process performance
Oxidation reduction potential (ORP) was monitored con-
tinuously in anaerobic stabilization pond in order to ensure
the anaerobic condition in this study. The average of (ORP <
–246) confirmed the anaerobic condition in stabilization pond.
3.2.1. Carbon removal
Effects of different phenol concentrations and tempera-
tures of stabilization pond at different HRT (2, 5) days on
responses (TBOD5, TCOD, SBOD5, and SCOD removal) are
presented in Tables 3 and 4. The removal efficiency trend
of SBOD5, SCOD changed when the phenol concentration
rose from 200 to 400 mg/L, and also the temperature from
8 to 24°C, at HRT of 5 d (Figs. 1 and 2). The removal effi-
ciency of SBOD5, SCOD increased considerably by raising
temperature and declining phenol concentration. The result
showed that temperature was a more effective factor in
terms of carbon removal in stabilization pond compared
to phenol concentration. It must be mentioned that varia-
tion of removal efficiency was more apparent at high sys-
e
17.7
30.84
43.97
57.11
70.24
Phenol removal, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol,
mg/l
B: Temperature,
°c
d
11.19
22.72
34.25
45.78
57.32
Phosphate removal, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg
/l
B: Temperature, °c
c
8.27
16.93
25.5
34.25
42.91
Nitrat removal, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg/l
B: Temperature, °c
22.92
34.10
45.28
56.45
67.63
Rem.TBOD5, %
200
250
300
350
400
8
12
16
20
24
B: Temperature, °c
A: Conc. Phenol, mg/l
b
23.7
34.89
46.08
57.27
68.46
Rem. TCOD, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg/l
B: Temperature, °c
a
Fig. 1. 3D Surface plots for a: TCOD removal , b: TBOD5 removal, c: Nitrate removal, d: Phosphate removal and e: Phenol removal
in HRT: 2 days.

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
204
tem temperatures. It was found that SCOD, TCOD, TBOD5
and SBOD5 removal efficiency increased 28.55, 13.32, 30.08
and 30.12%, respectively, when temperature increased from
8 to 24°C at HRT of 2 d. However, the removal efficiency
of SCOD, TCOD, TBOD and SBOD grew 16.85, 30.23, 13.52
and 15.63%, respectively, when phenol concentration went
down from 400 to 200 mg/L. The maximum removal effi-
ciency of SCOD, TCOD, SBOD and TBOD reached 66.26,
68.95, 65.3 and 67.02%, respectively (HRT of 2 d, tem-
perature of 24°C and phenol concentration of 200 mg/L).
It was observed that removal efficiency of SCOD, TCOD,
SBOD and TBOD reached 27.36, 26.63, 26.61 and 28.38%,
respectively (HRT of 5 d, temperature of 8°C and phenol
concentration of 400 mg/L). It should be mentioned that
volumetric loading of stabilization pond in phenol concen-
tration of 200, 300 and 400 mg/l, were 118.55, 131.74 and
143.48 g/BOD5/m3·d, respectively, at temperature of 24°C
and 121.54, 136.01 and 148.12 g/BOD5/m3·d, respectively,
at temperature of 8°C that leads to low efficiency of pond
by raising phenol concentration and lowering system tem-
e
21.98
36.64
51.3
65.97
80.63
Phenol removal, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg/l
B: Tem
p
erature
,
°c
d
15.56
27.11
38.67
50.22
61.78
Phosphate removal, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg/l
B: Temperature,
°c
c
12.63
23.59
34.54
45.5
56.45
Nitrat removal, %
200
250
300
350
400
8
12
16
20
24
B: Temperature,
°c A: Conc.phenol,
mg/l
26.5
37.93
49.36
60.79
72.2
Rem.TBOD5, %
200
250
300
350
400
8.
12
16
20
24
A: Conc.phenol, mg/l
B: Temperature, °c
b
27.14
39.59
52.03
64.48
76.93
Rem. TCOD, %
200
250
300
350
400
8
12
16
20
24
A: Conc.phenol, mg/l
B: Temperature, °c
a
Fig. 2. 3D Surface plots for a: TCOD removal , b: TBOD5 removal, c: Nitrate removal, d: Phosphate removal and e: phenol removal
in HRT: 5 days.

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
205
perature. When HRT rose from 2 to 5 d, the removal effi-
ciency of SCOD, TCOD, SBOD5 and TBOD went up 11.7,
9.67, 7.02 and 8.83%, respectively (at 24°C and phenol con-
centration of 400 mg/L), although this removal efficiency
was not significant. Less removal efficiency was observed at
lower temperature of anaerobic stabilization pond. It can be
attributed to lower microorganism growth and slower deg-
radation rate of dissolved substance. Although vast major-
ity of biodegradable substances are removed by hydrolysis
in anaerobic digestion, most suspended organic solids are
converted into soluble intermediate compounds with
poor biodegradability by hydrolysis in anaerobic stabiliza-
tion pond [18]. The anaerobic stabilization pond achieved
higher COD removal efficiency than BOD5 removal which
is not used for municipal and industrial wastewater. It may
be attributed to multi-phase process of oil refinery waste
due to volatile characteristics of upper layer, and also sepa-
ration of some layers owing to hydrophobic characteristics
of oily wastewater that stays or precipitates in the column
inside the reactor. Another notable reason for degradation
of non-biodegradable compounds converted into biode-
gradable fragments such as Catechol, aldehydes and acidic
materials and is higher COD removal efficiency compared
to BOD5. It was in contrast with the study carried out by
[19]. As result of surplus dissolving of the biodegradable
compound (SCOD) in the depth of anaerobic stabilization
pond, the overall decline of TCOD was observed. Also the
anaerobic stabilization pond was found to be more capa-
ble for the treatment of oil refinery wastewater contain-
ing different concentrations of phenol compared with the
study conducted by Mahssen et al. It showed that removal
efficiency of COD and BOD5 reached 28.89% and 22.21%,
respectively [20]. Effect of temperature on biological reac-
tion rate is completely obvious. Several studies have proved
the appropriate performance of stabilization pond in warm
weather [21].
3.2.2. Nutrient removal
Almost since the mid-1970, the use of algae to mitigate
the nutrients N and P in wastewater treatment has been
considered for combatting the Eutrophication [22]. Microal-
gae are a very diverse group of photosynthetic organisms
that absorb N and P during their growth, then the gener-
ated biomass can be converted into energy or more raw
materials depending on appropriate processing. Microalgae
are able to adjust the nutrient based on surrounding con-
centration [23]. However, N concentration in microalgae
depends on P concentration. Accordingly, present work
evaluated the ratio of N/P. Generally, the ratio of N/P
varied between 6.69–9.12 and 7.04–12.93, respectively, in
influent and effluent that are presented in Tables 1 and 2.
In general, the investigation showed that P removal was
higher than N removal, leading to increase in N/P ratio in
effluent than influent of stabilization pond, (Tables 1, 2). it
was in contrast with Whitton et al., study that evaluated the
influence of microalgae’s N and P composition on waste-
water nutrient remediation, and the result showed that the
N/P ratio ranged from 2.03 to 15 from the third to tenth
day [24]. According, the ratio of N/P declined from 12.16
to 7.04 in influent of system while phenol concentration
increased from 200 to 400 mg/L, in HRT of 2 d and tem-
perature of 24°C. The maximum and minimum ratio of N/P
were observed at temperature of 8 and 24°C, respectively
and HRT of 2 d. In other words, the lowest phenol concen-
tration was observed at low system temperature ; therefore,
increase in temperature resulted in improvement in the P
removal efficiency. It can be attributed to mesophilic micro-
organisms and algae that provide the optimum condition
for their growth [24].
Phosphate removal efficiency ranged from 2% to
14.48% and from 4.08% to 8.97%, respectively by HRT of
2, 5 d that was in contrast with Whitton el al. study. There-
fore, N concentration is considered as a limiting agent
in terms of raising the removal efficiency of stabilization
pond. Overall, the phosphate removal is associated with N
removal in cellular metabolism [25]. Mostly, in microalgae,
N integrated into protein for ribosome production and
ribosomal RNA (rRNA) while most P uptook is stored in
rRNA. Therefore,enough N is needed to ensure no limita-
tion for synthesis of protein. Based on studies carried out,
in low N environment, the P uptake into biomass remains
low, regardless of the P concentration in the biomass [26].
In another study, it was revealed that the uptake rate is
associated with the ability of microalgae to store available
phosphate via luxurious uptake pathway which accumu-
lates the polyphosphate within the cell [1]. In general,
based on the results of experiments a similar trend was
observed in terms of N and P removal efficiency at HRT
2, 5 d. Hence, increasing phenol concentration and lower-
ing temperature led to decrease in the removal efficiency
trend. As it is seen in Table 3, the coefficient of (A, B) in N
removal were 4.22 and 13.1, respectively, this means that
temperature has more positive effect than phenol concen-
tration, and the same trend for phenol removal efficiency
was observed. It should be taken into account that the dif-
ferent effects of microalgae in terms of nutrient removal
depends on the type of wastewater, wastewater treatment
technology, temperature and the condition of operation
system [27]. Orumieh et al. [27] indicated that maximum
removal efficiency of N and P were 33% and 25%, respec-
tively in stabilization pond. But the nutrient removal
efficiency was more satisfying in this study compared to
Orumieh’s study result. Generally, the experiment result
revealed that temperature variation was a more effective
factor in terms of nutrient removal in stabilization pond.
N concentration decreased due to cellular uptake at high
system temperature [27].
3.2.3. Phenol removal
In regard with transforming the waste into simple end
product, the use of biological treatment is increasing now-
adays [5]. The surface plot for phenol removal in anaerobic
stabilization pond is depicted in (Figs. 1 and 2), indicat-
ing the interaction effect of phenol concentration and
temperature. As it can be seen, raising temperature and
lowering phenol concentration enhanced phenol removal
efficiency. For instance, removal efficiency decreased from
20.51% to 14.67% and phenol concentration rose from 200
to 400 mg/L, at the same temperature of 24°C at HRT 2,
5 d, respectively. The result showed that phenol removal
efficiency varied between 18.22–70.53% and 21.24–81.63%,
respectively, at HRT 2 and 5 d. The maximum removal

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
206
efficiency of phenol concentration was observed in phe-
nol concentration of 200 mg/l and temperature of 24°C.
It was found that the lowest removal efficiency was
observed in maximum phenol concentration and lowest
system temperature. It was evident that phenol removal
efficiency was more affected by temperature than phenol
concentration. In consequence, the anaerobic digestion of
organic compounds such as phenol, produced the nutrient
rich sludge which can be used as fertilizer in agricultural
soil with low level of pathogen. Azbar et al. studied the
phenol removal in anaerobic hybrid reactor. The result
revealed that phenol removal efficiency ranged from 39
to 80% under different conditions that is consistent with
the study by [28]. Shabita et al. evaluated the microbial
degradation of various phenols and their derivations. In
this condition 4-n- propylphenol and phenol are degraded
[29]. In the first stepwise of anaerobic pathway, phenol is
carboxylated in the para position to 4-hydroxybenzonate,
the enzyme involved was the 4-hydrobenzonate carbox-
ylase [5]. This study represented that removal efficiency
increased slightly when HRT went up. The maximum
removal efficiency were 70.53% and 81.63 (at phenol con-
centration of 200 mg/L, temperature of 24°C and HRT of
2 and 5 d, respectively. The phenol removal efficiency was
more significantly influenced by high temperature than
decrease in the phenol concentration. According to Table 2,
it was found that the phenol removal efficiency enhanced
39.14% (Runs 8, 11) when temperature increased from 8 to
24°C, and phenol concentration declined from 400 to 200
mg/L at HRT of 2 d. And also phenol removal efficiency
increased 14.67%, (runs 1, 11). The effect of increase in tem-
perature and decrease in phenol concentration at HRT of 5
days was significantly apparent (Table 2).
3.3. Process optimization and verification
The experiment was carried out based on CCD to attain
process optimization. In this study, 3 independent variables
of phenol concentration, temperature and HRT were eval-
uated aiming at processing the optimization of the anaer-
obic stabilization pond. Graphical optimization produces
an overlay plot expressing the feasibility response value in
the factor space. Overlay plot represents the region which
meets the proposed criteria. The optimum area relative to 7
responses was determined (Fig. 3). The yellow area shows
the region that satisfies the responses. And the shaded
region is related to variables of space. According to the
response, the optimum region in terms of carbon, nutrient
and phenol removal were identified 60%, 40% and 70%,
respectively at HRT of 2 d and in terms of total carbon, sol-
uble carbon, nitrate, phosphate and phenol removal were
70%, 60%, 50%, 60% and 80%, respectively, at HRT of 5 d.
Aimed at confirming the adequacy and reliability of the
model, a point among optimum area was selected which
is depicted in Fig. 3 and the actual and predicted values of
model were compared. Table 5 shows the results of exper-
iments within the optimum region. The accuracy of opti-
mum condition was determined for each response from the
DOE experiments that was tested through applying stan-
dard deviation. The results revealed that experimental find-
ings were very close to predicted values by the model.
4. Conclusion
The results of experiment revealed raising phenol con-
centration led to decrease in the performance of anaerobic
stabilization pond due to increasing the phenol toxicity on
purifying bacteria in oil refinery wastewater treatment.
In order to analyze the resulting data, RSM was used to
demonstrate the effect of operating variables and interaction
effect on the response as well. The results demonstrated the
significant effect of HRT, temperature and phenol concen-
tration on anaerobic stabilization pond efficiency in pilot
scale. Temperature was more effective on Phenol removal
efficiency compared with phenol concentration. In general,
anaerobic stabilization pond was found to be a success-
ful biological treatment process for moderate removal of
organic and inorganic compounds using appropriate oper-
ations. The applicable properties of pond stabilization such
as flexibility, simplicity, performance and also operation
A: Phenol concentration, mg/l
B: Temperature, ◦c
200 250 300 350
400
8
12
16
20
24
5 5 5 5 5
X 300
Y16
b
A: Phenol concentration, mg/l
B: Temperature, ◦c
200 250 300 350 400
8
12
16
20
24
5 5 5 5 5
X 300
Y 16
a
Fig. 3. Overlay plots for the optimal region, a: HRT: 2 days, b: HRT: 5 days.

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
207
led to applying stabilization pond rather than complex and
expensive wastewater treatment technology such as acti-
vated sludge, etc.
References
[1] L. Wang, M. Min, Y. Li, P. Chen, Y. Chen, Y. Liu, Y. Wang, R.
Ruan, Cultivation of green algae Chlorella sp. in different
wastewaters from municipal wastewater treatment plant,
Appl. Biochem. Biotechnol., 162 (2010) 1174–1186.
[2] T. Misiti, U. Tezel, S.G. Pavlostathis, Fate and effect of naph-
thenic acids on oil refinery activated sludge wastewater treat-
ment systems, Water Res., 47 (2013) 449–460.
[3] Y.A. Yavuz, S. Koparal, U. Öğütveren, Treatment of petroleum
refinery wastewater by electrochemical methods, Desalina-
tion, 258 (2010) 201–205.
[4] V. Naddeo, A. Cesaro, D. mantzavinos, D. Fatta-kassinos, V.
Belgiorno, Water and wastewater disinfection by ultrasond
irradiation - A critical review, Glob NEST J., 16 (2014) 561–577.
[5] N.V. Pradeep, S. Anupama, J.M. Arunkumar, K.G. Vidyashree,
K. Ankitha, P. Lakshmi, J. Pooja, Treatment of sugar indus-
try wastewater in anaerobic down flow stationary fixed film
(DSFF) reactor., Sugar Tech, 16 (2014) 1–9.
[6] A. Almasi, A. Dargahi, A. Amrane, M. Fazlzadeh, M. Mah-
moudi, A.H. Hashemian, Effect of the retention time and the
phenol concentration the stabilization pond efficiency in the
treatment of oil refinery wastewater, FEB., 23 (2014) 2541–2548.
[7] A. Dargahi, A. Almasi, M. Soltanian, P. Zarei, A. Hashemian,
H. Golestanifar, Effect of molasses on phenol removal rate
using pilot-scale anaerobic reactors, J. Water Wastewater., 25
(2014) 2–12.
[8] Z. Cheng, F. Fu, Y. Pang, B. Tang, J. Lu, Removal of phenol by
acid-washed zero-valentaluminium in the presence of H2O2,
Chem. Eng. J., 260 (2015) 284–290.
[9] A. Moslehyani, A.F. Ismail, M.H.D Othman, T. Matsuura,
Design and performance study of hybrid photocatalytic reac-
tor-PVDF/MWCNT nanocomposite membrane system for
treatment of petroleum refinery wastewater, Desalination, 363
(2015) 99–111.
[10] S.O. Rastegar, S.M. Mousavi, S.A. Shojaosadatia, S. Sheibani,
Optimization of petroleum refinery effluent treatment in a
UASB reactor using response surface methodology, J. Hazard.
Mater., 197 (2011) 26– 32.
[11] A. Almasi, A. Dargahi, M.H. Ahagh, H. janjani, M. Moham-
madi, L. Tabandeh, Efficiency of a constructed wetland in con-
trolling organic pollutants, nitrogen, and heavy metals from
sewage, J. Chem. Pharm. Sci., 9 (2016) 2924–2928.
[12] N. Mburu, S.M. Tebitendwa, J.J.A. Van Bruggen, D.P.L Rous-
seau, P.N.L. Lens, Performance comparison and economics
analysis of waste stabilization ponds and horizontal subsur-
face flow constructed wetlands treating domestic wastewater:
A case study of the Juja sewage treatment works, J. Environ.
Manage., 128 (2013) 220–225.
[13] V. Del Nery, M.H.Z. Damianovic, E. Pozzi, I.R. de Nardi, V.E.A.
Caldas, E.C. Pires, Long-term performance and operational
strategies of a pou ltry slaughterhous e waste stabil ization pond
system in a tropical climate, J. Resour. Converserv. Recy., 71
(2013 ) 7–14.
[14] A. Almasi, A. Dargahi, M. Mohamadi, H. Biglari, F. Amirian,
M. Raei, Removal of Penicillin G by combination of sonolysis
and Photocatalytic (sonophotocatalytic) process from aqueous
solution: process optimization using RSM (Response Surface
Methodology). Electron Physician., 8 (2016) 2878–2887.
[15] E. Ghasemi, M. Sillanpää, Optimization of headspace solid
phase micro extraction based on nano-structured ZnO com-
bined with gas chromatography-mass spectrometry for pre
concentration and determination of ultra-traces of chloro
benzenes in environmental samples, Talanta, 130 (2014)
322–327.
[16] W. Apha, AWWA, Standard Methods for the Examination of
Water and Wastewater, 20th ed. American Public Health Asso-
ciation, Washington, DC, 2008.
[17] A.M. Antonopoulou, V. Papadopoulos, I. Konstantinou,
Photocatalytic oxidation of treated municipal wastewaters
for the removal of phenolic compounds: optimization and
modeling using response surface methodology (RSM) and
artificial neural networks (ANNs), J. Chem. Technol., 87
(2012) 1385 –1395.
[18] A. Almasi, A. Dargahi, M. Pirsaheb, The effect of different
concentrations of phenol on anaerobic stabilization pond per-
formance in treating petroleum refinery wastewater, J. Water
Wastewater, 1 (2013) 61–68.
[19] M. Pirsaheb, M. Mohamadi, A.M. Mansouri, A.A.L. Zinatiza-
deh, S. Sumathi, K. Sharafi, Process modeling and optimiza-
tion of biological removal of carbon, nitrogen and phosphorus
from hospital wastewater in a continuous feeding and inter-
mittent discharge (CFID) bioreactor, Korean J. Chem. Eng., 32
(2015) 1340 –1353.
Table 5
Verification experiments at optimum conditions
HRT,
d
Conditions Responses, %
TCOD
removal
SCOD
removal
TBOD5
removal
SBOD5
removal
N-NH3
removal
P-PO4
removal
Phenol
removal
2 Experimental
values
Phenol
concentration:
300 mg/l
Tem p erat ure,
16°C
46.0862 42.2354 43.8938 43.4838 22.6317 30.4192 42.5223
Model
response with
Cl 95% Error
45.04 40.95 42.47 42.42 21.13 29.09 41.02
Standard
deviation
1.75 2.15 2.40 1.78 1.66 2.20 2.52
5 Experimental
values
Phenol
concentration:
300 mg/l
Tem p erat ure,
16°C
52.7903 50.6654 51.3452 49.4631 30.1962 34.5738 45.791
Model
response with
Cl 95% Error
51.83 49.48 49.85 48.29 28.84 32.74 44.58
Standard
deviation
1.06 1.99 1.65 1.97 1.49 2.03 1.34

A. Dargahi et al. / Desalination and Water Treatment 87 (2017) 199–208
208
[20] A.M. Abdel-Aatty, M. Karnel, Performance evaluation of a
waste stabilization pond in a rural area in Egypt, Am. J. Envi-
ron. Sci., 4 (2008) 316–325.
[21] M.A.I. Al-Hashimi, H.T. Hussain, Stabilization pond for waste-
water treatment, Eur. Sci. J., 9 (2013) 67–74.
[22] H. Kocyig, A. Ugurlu, Biologycal decolorization of rective Azo
dye anarobic/aerobic sequencing batch reactor system, Glob.
NEST J., 17 (2015) 210–215.
[23] H.J. Choi, S.M. Lee, Effect of the N/P ratio on biomass produc-
tivity and nutrient removal from municipal wastewater, Bio-
process Biosyst. Eng., 38 (2015) 761–766.
[24] R. Whitton, A.L. Evel, P. Marc, F. Ometto, R. Villa, B. Jefferson,
Influence of microalgal N and P composition on wastewater
nutrient remediation, Water Res., 91 (2016) 371–378.
[25] I. Loladze, J.J. Elser, The origins of the Redfield nitro-
gen-to-phosphorus ratio are in a homoeostatic protein-to-
rRNA ratio, Ecol. Lett., 14 (2011) 244–250.
[26] A. Beuckels, E. Smolders, K. Muylaert, Nitrogen availability
influences phosphorus removal in microalgae-based wastewa-
ter treatment, Water Res., 77 (2015) 98–106.
[27] H.R Orumieh, R. Mazaheri, Efficiency of stabilization ponds
under different climate condition in Iran, Indian J. Fundam.
Appl. Life Sci., 5 (2015) 794–805.
[28] N. Azbar, F. Tutuk, T. Keskin, Biodegradation performance of
an anaerobic hybrid reactor treating olive mill effluent under
various organic loading rates, Int. Biodeterior. Biodegrad., 63
(2009) 690–698.
[29] A. Shibata, Y. Inoue, A. Katayama, Aerobic and anaerobic bio-
degradation of phenol derivatives in various paddy soils, Sci.
Total Environ., 367 (2006) 979–987.