Abstract

The purpose of this study was to optimize the coagulation–flocculation effect of a wastewater treatment system using the response surface methodology (RSM) and three-step method to minimize phosphorus concentration in the distillate wastewater. In order to minimize the concentration of total phosphorus (TP), experiments were carried out using -factorial designs with three levels and three factors. A Box–Behnken design, which is the standard design of RSM, was used to evaluate the effects and interactions of three major factors (Fe:P (w/w) ratio, coagulation pH and fast mixing speed (FMS)) on the treatment efficiency. A multivariable quadratic model developed for studying the response indicated that the values for optimum conditions for Fe:P (w/w) ratio, coagulation pH and FMS were 2.40, 6.48 and 100 rev min−1, respectively. Under optimal process conditions, the TP concentration in the distillery effluent was reduced from 10 mg L−1 to 0.215 mg L−1, representing a removal efficiency of 97.85%. Based upon the statistical evaluation of results, it is inferred that RSM can be used as an appropriate approach to optimize the coag-flocculation process. Meanwhile, the study has shown that, for the equivalent dose of ferric chloride, the average three-step effect is better than that of the one-time addition.

INTRODUCTION

Chinese liquor is one of the most important parts of traditional Chinese wine culture. With rapid social development and improved living standards, the demand for liquor has increased. Therefore, a lot of distillery effluent is being generated by distilleries. Distillery effluent is characterized by high concentration of suspended solids, chemical oxygen demand and phosphorus (Billore et al. 2001; Moletta 2005; Nataraj et al. 2006). Most of the distillery effluents are usually discharged directly without treatment, and lead to environmental pollution, which includes contamination of land and water resources. The high-phosphorus distillery effluent entering into the natural water body is significantly involved in eutrophication, which allows the algae to multiply, and results in microbial overgrowth and dissolved oxygen consumption (Correll 1998; Smith et al. 1999). In addition, the proliferation of large numbers of cyanobacteria may pose a potential hazard to humans and animals, as some of the cyanobacteria produce and release toxins, which are called cyanobacterial toxins (Codd et al. 1999; Dao et al. 2016). Therefore, it is necessary to remove phosphorus from distillery effluent before discharge. At present, in high-phosphorus wastewaters, phosphorus is removed mainly by a combination of chemical and biological methods. However, with this simple treatment it is difficult to meet the most stringent water discharge requirements. Therefore, it is urgent to optimize the chemical process in high-phosphorus wastewater treatment to remove phosphorus.

The coag-flocculation method has been widely used for removing phosphorus from wastewaters due to its advantages of simple operation, high efficiency and low cost. Iron and aluminum salts are the phosphorus removing agents, which are commonly used in coag-flocculation processes (Ebeling et al. 2003; Verma et al. 2012; Irfan et al. 2013; Kim et al. 2015). Compared with alum, iron salts are more promising for practical application to remove phosphorus because of their relatively low cost and high phosphorus removal efficiency (Huang et al. 2016). The process of treating industrial or agricultural high-phosphorus wastewater is accomplished by using the coag-flocculation method as shown in Figure 1.

Figure 1

Phosphorus-containing wastewater treatment process.

Figure 1

Phosphorus-containing wastewater treatment process.

Previous studies have demonstrated that ferrous or ferric salts are commonly used as flocculants for phosphorus removal in wastewater treatment (Priyantha & Perera 2000; Zhang et al. 2015; Wang et al. 2016). Seida et al. (2002) successfully synthesized iron-based layered double hydroxides which removed more than 80% of the phosphate from wastewater. Ivanov et al. (2009) studied the reduction of iron ore using iron-reducing bacteria to remove phosphate from sewage water and obtained a phosphorus removal efficiency of about 90%.

The optimization of the coagulant dosage is of great significance to improve the efficiency of phosphorus removal and reduce the operational cost. Considering both the influences of individual factors and their interactions, response surface methodology (RSM) has been proposed as a potential solution. RSM is an empirical statistical modeling technique used to design experiments, build models, and find the optimum conditions for an ideal response, while involving a limited number of planned experiments (Gulati et al. 2010). Recently, RSM has also been used to study the optimization of coagulant dosage and flocculant in the coag-flocculation process (Khayet et al. 2011; Trinh & Kang 2011; Agbovi & Wilson 2017; Momeni et al. 2018).

In order to reduce the cost of phosphorus removal, this study uses RSM or the three-step method to optimize the coag-flocculation process for phosphorus removal in treating high-phosphorus wastewater. The study was conducted with the following objectives: (1) to evaluate the feasibility of ferric chloride coagulant for removing phosphorus from distillery effluent; (2) to design the coag-flocculation experiments using Box–Behnken design and optimize the coagulant dosage using RSM; (3) to design the three-step experiments to reduce the cost of phosphorus removal.

MATERIALS AND METHODS

Materials and water samples

All reagents used in the experiments were of analytical reagent grade and used without further purification. Ferric chloride hexahedron (FeCl3•6H2O, 99.0%) and HCl (35–37%) were purchased from Tianjin Sailboat Chemical Reagent Technology Co. Ltd (China), while sodium hydroxide (NaOH, 97%) was obtained from Tianjin Chemical Reagent Factory (China). Furthermore, sodium bicarbonate (NaHCO3, 99.9%) was purchased from Tianjin Yongda Chemical Reagent Co. Ltd (China). The concentration of phosphorus in the distillery effluent was 10 mg P L−1.

Phosphorus removal experiments

Various factors affecting the removal of phosphorus using ferric chloride include Fe:P (w/w) ratio, fast mixing time (FMT), fast mixing speed (FMS) and pH. These parameters were investigated using batch experiments. All experiments were performed in triplicate. The phosphorus removal rate was calculated using Equation (1): 
formula
(1)
where Co and Ce are the initial and equilibrium phosphorus concentrations in mg L−1, respectively.

Effect of Fe:P (w/w) ratio

Batch experiments were conducted using different volumes of 10 g L−1 FeCl3. These values were 1, 2, 3, 4, 5 and 6 mL. Each FeCl3 solution was added into a 500 mL beaker containing 400 mL of distillery effluent (10 mg P L−1), and was then mixed using a magnetic stirrer (IKA, RO5, Germany) at 200 rev min−1 and 25 °C for 5 min. After the mixture was permitted to stand for 40 min, 25 mL of the supernatant was withdrawn and then subjected to total phosphorus concentration measurement using ammonium molybdate spectrophotometry (absorbed light at 700 nm) method using a UV-Vis spectrophotometer (UV-2550, Hach Co., USA). The performance of ferric chloride for phosphorus removal was then evaluated.

Effect of fast mixing time

In order to investigate the effect of FMT on the removal of phosphorus using ferric chloride, batch experiments were carried out at six different FMTs (2, 3, 4, 5, 6 and 7 min), and using 400 mL of distillery effluent (10 mg P L−1).

Effect of fast mixing speed

In order to investigate the effect of FMS on the removal of phosphorus, batch phosphorus removal experiments were carried out at five different FMSs (150, 200, 250, 300 and 350 rev min−1), using 400 mL of distillery effluent (10 mg P L−1) for each FMS.

Effect of pH

The experiment was designed to investigate the effects of different pH values (with the range of 5–10) on the removal of phosphorus using ferric chloride. Four hundred millilitres of distillery effluent (10 mg P L−1) with different initial pH values was used in the batch experiments.

Optimization using Box–Behnken design

Box–Behnken experimental design with three levels and three factors combined with RSM (Beker et al. 2013), is a collection of mathematical and statistical techniques for designing experiments. The design is a useful approach that can be utilized to study the effect of various factors influencing the responses, in which the factors can be simultaneously varied during a small number of experiments (Adinarayana & Ellaiah 2002). Since FMT has negligible effect on the phosphorus removal using ferric chloride, this factor was not considered in RSM. Other factors included Fe:P (w/w) ratio, pH and FMS, and were assigned as independent factors. The efficiency of ferric chloride for phosphorus removal at equilibrium was assigned as the response to the modified process (dependent variable). The values of these factors were for three levels of −1, 0 and +1, which represented low (Zhang et al. 2011), central and high values, respectively (Table 1). The values of Fe:P (w/w) ratio (X1), pH (X2) and FMS (X3) were set as the input factors. Table 2 shows the coded levels of factors that were employed in these experiments.

Table 1

Response surface factor levels

No. Independent factors Code Levels
 
− 1 
Fe:P(w/w) 2.0 2.2 2.4 
Fast mixing speed 100 150 200 
pH 
No. Independent factors Code Levels
 
− 1 
Fe:P(w/w) 2.0 2.2 2.4 
Fast mixing speed 100 150 200 
pH 
Table 2

Experimental design and results of response surface

Run Factor 1 A: Fe:P (w/w/) Factor 2 B: FMS (rev min−1Factor 3 C: pH %Removal 
2.0 100 87.75 
2.4 100 94.50 
2.0 200 80.25 
2.4 200 86.00 
2.0 150 56.25 
2.4 150 73.00 
2.0 150 78.75 
2.4 150 73.00 
2.2 100 85.00 
10 2.2 200 47.75 
11 2.2 100 61.00 
12 2.2 200 89.50 
13 2.2 150 86.25 
14 2.2 150 85.50 
15 2.2 150 86.57 
16 2.2 150 89.50 
17 2.2 150 86.25 
Run Factor 1 A: Fe:P (w/w/) Factor 2 B: FMS (rev min−1Factor 3 C: pH %Removal 
2.0 100 87.75 
2.4 100 94.50 
2.0 200 80.25 
2.4 200 86.00 
2.0 150 56.25 
2.4 150 73.00 
2.0 150 78.75 
2.4 150 73.00 
2.2 100 85.00 
10 2.2 200 47.75 
11 2.2 100 61.00 
12 2.2 200 89.50 
13 2.2 150 86.25 
14 2.2 150 85.50 
15 2.2 150 86.57 
16 2.2 150 89.50 
17 2.2 150 86.25 
Design-Expert 10 (Stat-Ease Inc., USA) was used to design the experiments. The sequential model-fitting test was carried out to find a suitable model. The Box–Behnken model, which is composed of a second-order polynomial model, was identified, and all possible interactions between the selected factors were expressed using Equation (2). 
formula
(2)

Simulating the three-step method to remove phosphorus from distillery effluent

First, 400 mL of distillery effluent was taken in a beaker. Then 0.8 mL of 10 g L−1 ferric chloride solution was added into the beaker, and mixed using a magnetic stirrer (IKA, RO5, Germany) at 200 rev min−1 and 25 °C for 30 min. After the mixture was allowed to stand overnight, 25 mL of the supernatant was withdrawn and subjected to total phosphorus concentration measurement using ammonium molybdate spectrophotometry (absorbed light at 700 nm) method using a UV-Vis spectrophotometer (UV-2550, Hach Co., USA). The experiments were performed three times to ensure the repeatability of results.

RESULTS AND DISCUSSION

Phosphorus removal experiments

Effect of Fe:P (w/w)

Batch experiments were performed to determine the optimal of Fe:P (w/w) ratio for phosphorus removal. According to Figure 2(a), the phosphorus removal rates rapidly increased with the increase in Fe:P (w/w) ratio from 0.4 to 2.15. In contrast, when the Fe:P (w/w) ratios were greater than 2.15, the phosphorus removal rate remained nearly unchanged, indicating that the optimal Fe:P (w/w) ratio is at 2.15. For Fe:P (w/w) ratio of less than 2.15, with the increase in the amount of ferric chloride, the volumes of flocs increased due to the increasing content of absorbed phosphorus, which then induced further complexation on the surface of flocs. As a result, the phosphorus anions in water were further reduced. The process can be explained using the reactions expressed in Equations (3)–(6) (Lijklema 1980; Smith et al. 2008; Szabó et al. 2008). On the other hand, when Fe:P (w/w) ratio was greater than 2.15, the hydrolysis of ferric chloride may be inhibited due to high acidity in water, which was caused by ferric chloride. As a result, flocs no longer formed, and thus the rate of phosphorus removal did not show significant differences. 
formula
(3)
 
formula
(4)
 
formula
(5)
 
formula
(6)
Figure 2

Effect of individual factors on phosphorus removal. (a) Effect of dosage of ferric chloride; (b) effect of fast mixing time; (c) effect of fast mixing speed; (d) effect of pH.

Figure 2

Effect of individual factors on phosphorus removal. (a) Effect of dosage of ferric chloride; (b) effect of fast mixing time; (c) effect of fast mixing speed; (d) effect of pH.

Effect of fast mixing time

An appropriate FMT allows better dissolution and spread of ferric chloride in water, and also increases the possibility of contact between the phosphorus and flocs, thus enhancing the removal of phosphorus. In order to examine the effect of FMT on the efficiency of ferric chloride in phosphorus removal, the phosphorus removal experiment was conducted by changing the FMT within the range of 2–7 min. The corresponding results are shown in Figure 2(b), which shows that with the increase of FMT, the phosphorus removal slightly fluctuated between the values of 76% and 85%, and reached a small peak at 6 min. It appears that the effect of FMT on the phosphorus removal is insignificant and can be ignored. Therefore, it was not considered in the optimization of RSM using the Box–Behnken experimental design.

Effect of fast mixing speed

At a certain appropriate FMS, the rate of phosphorus removal using ferric chloride can be increased because the solution is more homogeneous, which enhances the probability of contact between the hydrated iron complex ions and the phosphate anion. In order to determine the optimum value for FMS, the phosphorus removal was performed at various FMS values, within the range of 150–350 rev min−1. The results shown in Figure 2(c) demonstrate that the removal of phosphorus increased from 77.00% to 84.50% with the increase in FMS from 150 to 200 rev min−1. The phosphorus removal is optimum at the FMS of 200 rev min−1, after which the effect of FMS on phosphorus removal becomes insignificant. Therefore, the results indicate that the optimum FMS is 200 rev min−1 for achieving the highest efficiency of phosphorus removal.

Effect of pH

The quantity of flocs formed during phosphorus removal varies with pH (Lijklema 1980; Zinatizadeh et al. 2006; Nie et al. 2015). Therefore, phosphorus removal at different pH, ranging from 5.0 to 10.0, was examined to study the effect of pH. Figure 2(d) illustrates the effect of solution pH on phosphate removal by ferric chloride. It is generally known that the negatively charged phosphate (usually H2PO4 or HPO42−) is selectively trapped by flocs through formation of the inner-sphere complexes, and HPO42− tends to form stronger bidentate complexes than H2PO4 (Zeng et al. 2008). As shown in Figure 2(d), as the pH is from 5 to 7, the H2PO4 in the water is converted to HPO42−, so the efficiency of phosphorus removal is increased. However, in alkaline conditions, the negatively charged sites on the surface of flocs became dominant and the negatively charged phosphate species increased (Cumbal & Sengupta 2005), which resulted in the repulsion between the phosphate ions and the flocs. As a result, phosphate adsorption dropped remarkably.

When the pH increased from 5.0 to 7.0, the phosphorus removal increased from 19.25% to 89.25%. However, the phosphorus removal quickly declined from 89.25% to 69.50% when the pH was increased from 7.0 to 10.0. Therefore, the optimal pH for the removal of phosphorus using ferric chloride was found to be 7.0.

Analysis of variance

Table 3 presents the results of the analysis of variance (ANOVA) of the second-order polynomial equations and the corresponding regression coefficients. The analysis showed that the model had an F value of 114.19 with Prob far greater than F of 0.0001, which indicates that the model is significant and partly practical. Adequate precision compares a range of the predicted values of design space, and in turn, is desirable (Mohajeri 2010). In the present work, adequate precision of 37.983 indicates that the model is acceptable. Additionally, the experimental R2 value of 0.9932 is comparable to the model's R2 value of 0.9845, which indicates that the model is in good agreement with the experimental data (Adinarayana & Ellaiah 2002). Moreover such an R2 value indicates that the model can explain about 99% of the experimental data. The significant lack of fit observed in this study is suitable for the model. Depending on the results, the response surface model constructed for predicting the efficiency of ferric chloride in phosphorus removal can be considered accurate. The final regression model is based on the defined factors, A, B and C, which can be expressed using Equation (7). 
formula
(7)
Suitability of the model, which provides adequate approximation of the real system, is confirmed based upon the diagnostic plots. Such plots, which include normal probability plots of studentized residuals and the plot of predicted value versus actual value, are used to judge the adequacy of a model. Figure 3(a) shows the normal probability plot of the studentized residuals in phosphorus removal. Studentized residuals represent a normal probability plot, in which the residuals follow a normal distribution, while the points follow a straight line, and some scattered data points can generally be expected. The data are evenly distributed (Figure 3(a)), and the predicted values are in good agreement with the experimental values (Figure 3(b)).
Table 3

ANOVA for phosphorus removal

Source Sum of squares df Mean square F Value Prob > F Comment 
Model 2,717.85 301.98 114.19 <0.0001 SD = 1.63 
A: Fe:P 69.03 69.03 26.10 0.0014 Mean = 79.22 
B: speed 78.13 78.13 29.54 0.0010 CV = 2.05 
C: pH 205.03 205.03 77.53 <0.0001 Press = 157.96 
AB 0.25 0.25 0.095 0.7674 R2 = 0.9932 
AC 126.56 126.56 47.86 0.0002 R2(adj) = 0.9845 
BC 1,089.00 1,089.00 411.78 <0.0001 AP = 37.983 
A2 0.053 0.053 0.020 0.8911  
B2 0.63 0.63 0.24 0.6398  
C2 1,144.58 1,144.58 432.79 <0.0001  
Residual 18.51 2.64    
Lack of fit 8.94 2.98 1.24 0.4043  
Pure error 9.57 2.39    
Cor total 2,736.36 16     
Source Sum of squares df Mean square F Value Prob > F Comment 
Model 2,717.85 301.98 114.19 <0.0001 SD = 1.63 
A: Fe:P 69.03 69.03 26.10 0.0014 Mean = 79.22 
B: speed 78.13 78.13 29.54 0.0010 CV = 2.05 
C: pH 205.03 205.03 77.53 <0.0001 Press = 157.96 
AB 0.25 0.25 0.095 0.7674 R2 = 0.9932 
AC 126.56 126.56 47.86 0.0002 R2(adj) = 0.9845 
BC 1,089.00 1,089.00 411.78 <0.0001 AP = 37.983 
A2 0.053 0.053 0.020 0.8911  
B2 0.63 0.63 0.24 0.6398  
C2 1,144.58 1,144.58 432.79 <0.0001  
Residual 18.51 2.64    
Lack of fit 8.94 2.98 1.24 0.4043  
Pure error 9.57 2.39    
Cor total 2,736.36 16     

df, degrees of freedom; SD, standard deviation; CV, coefficient of variation; Press, predicted residual error sum of squares; AP, adequate precision; Cor total, totals of all information corrected for the mean.

Figure 3

(a) Normal plot of residuals; (b) plot of predicted v/s actual.

Figure 3

(a) Normal plot of residuals; (b) plot of predicted v/s actual.

Three-dimensional response surface plot

Wu et al. (2009) have reported that the three-dimensional (3D) response surface is a function of two factors, while all other factors remain unchanged. This helps in understanding the main effects as well as interaction effects of the two factors. In addition to the 3D response surface, the corresponding contour map can help visually verify the influence of experimental variables on the response (Szabó et al. 2008). In order to better understand the interaction effects between the flotation variables, a 3D response surface map of the measured response was generated based on Equation (2) (Figure 4(a)–4(f)). Moreover, contour plots can further clarify the relationship between the dependent and independent variables, while one variable in each graph remained unchanged at the center level. Therefore, three response 3D maps and three corresponding contour maps were generated.

Figure 4

Three-dimensional surface and two-dimensional contour plots of phosphorus removal. (a) and (b) Effect of Fe:P (w/w) and fast mixing speed; (c) and (d) effect of Fe:P (w/w) and pH; (e) and (f) effect of pH and fast mixing speed.

Figure 4

Three-dimensional surface and two-dimensional contour plots of phosphorus removal. (a) and (b) Effect of Fe:P (w/w) and fast mixing speed; (c) and (d) effect of Fe:P (w/w) and pH; (e) and (f) effect of pH and fast mixing speed.

Figure 4(a) and 4(b) show the two-dimensional contour map and the 3D response surface relationship between Fe:P (w/w) ratio (A) and FMS (B) at the center level of pH (C). The phosphorus removal increased with the increase in Fe:P (w/w) ratio (A). Similar trends were observed in the two-dimensional contour map and 3D response surface relationship between Fe:P (w/w) ratio (A) and pH (C) at the center level of FMS (B), which are shown in Figure 4(c) and 4(d). The efficiency for phosphorus removal initially increased and then decreased with the increase in pH value. Figure 4(e)–4(f) display the two-dimensional contour map and 3D response surface relationship between FMS (B) and pH (C) at the center level of Fe:P (w/w) ratio (A). Similarly, efficiency in phosphorus removal increases and then decreases with increasing pH.

Optimization of phosphorus removal conditions

The optimization of numerical conditions for phosphorus removal was performed using Design-Expert software. Each operational condition (Fe:P (w/w), pH and FMS) was chosen within the range at which the response (phosphorus removal) was maximized (indicating the highest performance). The software combines the desired unique value into a single number, and then maximizes its function. The model prediction shows that, for 97.85% removal of phosphorus, the Fe:P (w/w), pH and FMS values were 2.40, 6.48 and 100 rev min−1, respectively.

Simulation of the three-step method to remove phosphorus from distillery effluent

As can be seen from Figure 5, after every addition of 8 mg ferric chloride to 400 mL distillery effluent, the total phosphorus concentration was reduced, and the efficiency of phosphorus removal reached 99.0% after three repetitions. If ferric chloride is added only once to distillery effluent, and the efficiency of phosphorus removal of more than 90% is desired, the dosage of ferric chloride should be increased to 125 mg L−1. Compared with one-time dosage of ferric chloride, the use of the three-step method to treat distillery effluent can reduce its operational cost. Figure 6 shows the industrial process of phosphorus removal.

Figure 5

Using three-step method to remove phosphorus in distillery effluent.

Figure 5

Using three-step method to remove phosphorus in distillery effluent.

Figure 6

Simulating the industrial treatment of distillery wastewater using three-stage treatment.

Figure 6

Simulating the industrial treatment of distillery wastewater using three-stage treatment.

CONCLUSION

The presence of phosphorus in distillery effluent is one of the reasons for hindering its reuse for farming and drinking. The Coag-flocculation employed in this study is able to reduce phosphorus concentration below the World Health Organization permissible level of 0.5 mg L−1. The RSM was successfully applied to optimize the phosphorus removal from aqueous solution. At optimal operating conditions of Fe:P (w/w) ratio = 2.40, pH = 6.48 and FMS = 100 rev min1, 97.85% removal of phosphorus was achieved. Moreover, compared with the one-off dosage, the three-step process treatment can save 50% of the cost in terms of ferric chloride when treating the same amount of distillery effluent.

ACKNOWLEDGEMENTS

This work was financially supported by Scientific Research Foundation of Yunnan Provincial Department of Education (No. 2018JS005), Natural Science Foundation of Yunnan University (No. 2017YDQN01), Major project of Kunming Science and Technology bureau (No. 2017-1-S-12305) and Academician Free Inquiry Project of Yunnan Province Science and Technology Department (No. 2017HA005).

REFERENCES

REFERENCES
Adinarayana
K.
&
Ellaiah
P.
2002
Response surface optimization of the critical medium components for the production of alkaline protease by a newly isolated Bacillus sp
.
Journal of Pharmacy & Pharmaceutical Sciences
5
,
272
.
doi: 10.1002/pd.4103
.
Agbovi
H. K.
&
Wilson
L. D.
2017
Flocculation optimization of orthophosphate with FeCl3 and alginate using the Box–Behnken response surface methodology
.
Industrial & Engineering Chemistry Research
56
.
doi: 10.1021/acs.iecr.6b04765
.
Beker
U.
,
Tuna
A. Ö. A.
,
Özdemir
E.
&
Simsek
E. B.
2013
Optimization of process parameters for removal of arsenic using activated carbon-based iron-containing adsorbents by response surface methodology
.
Water Air and Soil Pollution
224
,
1685
.
doi: 10.1007/s11270-013-1685-z
.
Billore
S. K.
,
Singh
N.
,
Ram
H. K.
,
Sharma
J. K.
,
Singh
V. P.
,
Nelson
R. M.
&
Dass
P.
2001
Treatment of a molasses based distillery effluent in a constructed wetland in central India
.
Water Science & Technology
44
,
441
.
doi: 10.1016/S0043 −1354 (00) 00228-1
.
Codd
G.
,
Bell
S.
,
Kaya
K.
,
Ward
C.
,
Beattie
K.
&
Metcalf
J.
1999
Cyanobacterial toxins, exposure routes and human health
.
British Phycological Bulletin
34
,
11
.
doi: 10.1080/09670269910001736462
.
Correll
D. L.
1998
The role of phosphorus in the eutrophication of receiving waters: a review
.
Journal of Environmental Quality
27
,
261
266
.
doi: 10.2134/jep1998.00472425002700020004x
.
Cumbal
L.
&
Sengupta
A. K.
2005
Arsenic removal using polymer-supported hydrated iron(III) oxide nanoparticles: role of donnan membrane effect
.
Environmental Science & Technology
39
(
17
),
6508
6515
.
doi: 10.1021/es050175e
.
Dao
T. S.
,
Nimptsch
J.
&
Wiegand
C.
2016
Dynamics of cyanobacteria and cyanobacterial toxins and their correlation with environmental parameters in Tri An reservoir, Vietnam
.
Journal of Water and Health
14
,
699
712
.
doi: 10.2166/wh.2016.257
.
Ebeling
J. M.
,
Sibrell
P. L.
,
Ogden
S. R.
&
Summerfelt
S. T.
2003
Evaluation of chemical coagulation–flocculation aids for the removal of suspended solids and phosphorus from intensive recirculating aquaculture effluent discharge
.
Aquacultural Engineering
29
,
23
42
.
doi: 10.1016/s0144-8609(03)00029-3
.
Gulati
T.
,
Chakrabarti
M.
,
Sing
A.
,
Sing
A.
,
Duvuuri
M.
&
Banerjee
R.
2010
Comparative study of response surface methodology, artificial neural network and genetic algorithms for optimization of soybean hydration
.
Food Technology & Biotechnology
48
,
11
18
.
doi: 10.1080/08905436.2010.524489
.
Huang
S.
,
Huang
H.
&
Zhu
H.
2016
Effects of the addition of iron and aluminum salt on phosphorus adsorption in wetland sediment
.
Environmental Science and Pollution Research
23
(
10
),
10022
10027
.
doi: 10.1007/s11356-016-6188-1
.
Irfan
M.
,
Butt
T.
,
Imtiaz
N.
,
Abbas
N.
,
Khan
R. A.
&
Shafique
A.
2013
The removal of COD, TSS and colour of black liquor by coagulation–flocculation process at optimized pH, settling and dosing rate
.
Arabian Journal of Chemistry
.
S1878535213002682
.
doi: 10.1016/j.arabjc.2013.08.007
.
Ivanov
V.
,
Kuang
S.
,
Stabnikov
V.
&
Guo
C.
2009
The removal of phosphorus from reject water in a municipal wastewater treatment plant using iron ore
.
Journal of Chemical Technology & Biotechnology
84
,
5
.
doi:10.1002/jctb.2009
.
Khayet
M.
,
Zahrim
A. Y.
&
Hilal
N.
2011
Modelling and optimization of coagulation of highly concentrated industrial grade leather dye by response surface methodology
.
Chemical Engineering Journal
167
,
77
83
.
doi: 10.1016/j.cej.2010.11.108
.
Kim
W. K.
,
Sung
Y. K.
,
Yoo
H. S.
&
Kim
J. T.
2015
Optimization of coagulation/flocculation for phosphorus removal from activated sludge effluent discharge using an online charge analyzing system titrator (CAST)
.
Journal of Industrial and Engineering Chemistry
21
,
269
277
.
doi: 10.1016/j.jiec.2014.02.034
.
Lijklema
L.
1980
Interaction of orthophosphate with iron (III) and aluminum hydroxides
.
Environmental Science & Technology
14
,
537
541
.
doi: 10.1021/es60165a013
.
Moletta
R.
2005
Winery and distillery wastewater treatment by anaerobic digestion
.
Water Science & Technology
51
,
137
144
.
doi: 10. 1016/j.watres.2004.10.012
.
Momeni
M. M.
,
Kahforoushan
D.
,
Abbasi
F.
&
Ghanbarian
S.
2018
Using chitosan/CHPATC as coagulant to remove color and turbidity of industrial wastewater: optimization through RSM design
.
Journal of Environmental Management
211
,
347
355
.
doi: 10.1016/j.jenvman.2018.01.031
.
Nataraj
S. K.
,
Hosamani
K. M.
&
Aminabhavi
T. M.
2006
Distillery wastewater treatment by the membrane-based nanofiltration and reverse osmosis processes
.
Water Research
40
,
2349
2356
.
doi: 10.1016/j.watres.2006.04.022
.
Nie
G.
,
Wang
J.
,
Pan
B.
&
Lv
L.
2015
Surface chemistry of polymer-supported nano-hydrated ferric oxide for arsenic removal: effect of host pore structure
.
Science China Chemistry
58
,
722
730
.
doi:10.1007/s11426-014-5285-6
.
Priyantha
N.
&
Perera
S.
2000
Removal of sulfate, phosphate and colored substances in wastewater effluents using feldspar
.
Water Resources Management
14
,
417
434
.
doi: 10.1023/A:1011171330097
.
Seida
Y.
,
Nakano
Y.
&
Nakamura
Y.
2002
Crystallization of layered double hydroxides by ultrasound and the effect of crystal quality on their surface properties
.
Clays & Clay Minerals
50
,
525
532
.
doi: 10.1346/000986002320514244
.
Smith
V. H.
,
Tilman
G. D.
&
Nekola
J. C.
1999
Eutrophication: impacts of excess nutrient inputs on freshwater, marine, and terrestrial ecosystems
.
Environmental Pollution
100
,
179
196
.
doi: 10.1016/S0269-7491(99)00091-3
.
Smith
S.
,
Takács
I.
,
Murthy
S.
,
Daigger
G. T.
&
Szabó
A.
2008
Phosphate complexation model and its implications for chemical phosphorus removal
.
Water Environment Research
80
,
428
438
.
doi: info:doi/10.2175/106143008X268443
.
Szabó
A.
,
Takács
I.
,
Murthy
S.
,
Daigger
G. T.
&
Smith
S.
2008
Significance of design and operational variables in chemical phosphorus removal
.
Water Environment Research
80
,
407
416
.
doi:info:doi/10.2175/106143008X268498
.
Trinh
T. K.
&
Kang
L. S.
2011
Response surface methodological approach to optimize the coagulation–flocculation process in drinking water treatment
.
Chemical Engineering Research & Design
89
,
1126
1135
.
doi: 10.1016/j.cherd.2010.12.004
.
Verma
A. K.
,
Dash
R. R.
&
Bhunia
P.
2012
A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters
.
Journal of Environmental Management
93
,
154
168
.
doi: 10. 1016/j.jenvman.2011.09.012
.
Wang
L.
,
Hwang
J.
,
Xue
G.
&
Zhang
L.
2016
Preparation of polymeric phosphate ferric sulfate flocculant and application on coking wastewater treatment
.
Characterization of Minerals Metals & Materials
43
,
531
538
.
doi: 10.1007/987-3-319-48210-1-66
.
Wu
D.
,
Zhou
J.
&
Li
Y.
2009
Effect of the sulfidation process on the mechanical properties of a CoMoP/Al2O3 hydrotreating catalyst
.
Chemical Engineering Science
64
,
198
206
.
doi:10.1016/j.ces.2008.10.014
.
Zeng
H.
,
Fisher
B.
&
Giammar
D. E.
2008
Individual and competitive adsorption of arsenate and phosphate to a high-surface-area iron oxide-based sorbent
.
Environmental Science & Technology
42
(
1
),
147
152
.
doi:10.1021/es071553d
.
Zhang
H.
,
Ran
X.
,
Wu
X.
&
Zhang
D.
2011
Evaluation of electro-oxidation of biologically treated landfill leachate using response surface methodology
.
Journal of Hazardous Materials
188
,
261
268
.
doi:10.1016/j.jhazmat.2011.01.097
.
Zhang
Z.
,
Wang
Y.
,
Leslie
G. L.
&
Waite
T. D.
2015
Effect of ferric and ferrous iron addition on phosphorus removal and fouling in submerged membrane bioreactors
.
Water Research
69
,
210
222
.
doi: 10.1016/j.wartes.2014.11.011
.
Zinatizadeh
A. A. L.
,
Mohamed
A. R.
,
Abdullah
A. Z.
,
Mashitah
M. D.
,
Isa
M. H.
&
Najafpour
G. D.
2006
Process modeling and analysis of palm oil mill effluent treatment in an up-flow anaerobic sludge fixed film bioreactor using response surface methodology (RSM)
.
Water Research
40
,
3193
3208
.
doi:10.1016/j.watres.2006.07.005
.