This study deals with chemical oxygen demand (COD), phenol and Ca+2 removal from paper mill industry wastewater by electrocoagulation (EC) and electro-Fenton (EF) processes. A response surface methodology (RSM) approach was employed to evaluate the effects and interactions of the process variables and to optimize the performance of both processes. Significant quadratic polynomial models were obtained (R2 = 0.959, R2 = 0.993 and R2 = 0.969 for COD, phenol and Ca+2 removal, respectively, for EC and R2 = 0.936, R2 = 0.934 and R2 = 0.890 for COD, phenol and Ca+2 removal, respectively). Numerical optimization based on desirability function was employed; in a 27.55 min trial, 34.7% of COD removal was achieved at pH 9 and current density 96 mA/cm2 for EC, whereas in a 30 min trial, 74.31% of COD removal was achieved at pH 2 and current density 96 mA/cm2 and H2O2/COD molar ratio 2.0 for EF. The operating costs were calculated to be 6.44 €/m3 for EC and 7.02 €/m3 for EF depending on energy and electrode consumption at optimum conditions. The results indicate that the RSM is suitable for the design and optimization of both of the processes. However, EF process was a more effective technology for paper mill industry wastewater treatment as compared with EC.

High water consumption is one of the most important environmental concerns in the paper industry. Paper industry wastewater contains a large amount of pollutants characterized by biochemical oxygen demand, chemical oxygen demand (COD), total suspended solids, colorants, adsorbable organic halogens, phenols, heavy metals, sulphides and other soluble substances (Muhamad et al. 2013; Aghdam et al. 2016; Asaithambi et al. 2016; Chu et al. 2016; Un et al. 2016) and has a significant toxic effect on aquatic ecosystem and human health (Pokhrel & Viraraghavan 2004; Katal & Pahlavanzadeh 2011; Ashrafi et al. 2015; Jaafarzadeh et al. 2016). The main treatment processes used in paper industry wastewater treatment are primary clarification (sedimentation or flotation), secondary treatment (activated sludge process or anaerobic digestion) and/or tertiary processes (membrane processes as ultrafiltration, adsorption, ozone treatment and electro-chemical processes) (Thomson et al. 2001; Katal & Pahlavanzadeh 2011).

Aerobic biological processes are high energy-intensive processes, whereas anaerobic biological process effluents need additional treatment before final discharge (Kushwaha et al. 2010). Anaerobic processes are also prone to shock loading. In aerobic biological processes, large amounts of biomass are produced (Tchobanoglous et al. 2003). The activated sludge process fails to show good performance for treatment of liner paper wastewater since it contains high levels of organics and the high temperature of paper industry wastewater may cause serious settling problems such as bulking (Kim et al. 2003). There have been very few attempts to remove organics from wastewater containing high concentrations of calcium using upflow anaerobic sludge blanket reactors (Habets & Knelissen 1985; Matt & Habet 1987; Van Langerak et al. 1997). On the other hand, physical/chemical methods have the disadvantage of high reagent costs and low COD removal (Demirel et al. 2005; Tchamango et al. 2010). Chemical precipitation of dissolved matter by using chemicals (alum, ferric chloride, aluminum chloride, ferrous sulfate, and ammonium aluminum sulfate) has the disadvantage of dewatering and disposal of the precipitated sludge problems. Moreover, chemical treatment could induce a secondary pollution due to the chemical additives. In recent years, among different techniques used for industrial wastewater treatment, electrochemical methods have been widely applied.

Electrochemical technologies have received significant attention during the past two decades because they offer the possibility to be easily distributed and require minimum amounts and numbers of chemicals. Rajeshwar et al. (1994) listed the benefits of electrochemical techniques as environmental compatibility, versatility, energy efficiency, safety, selectivity, amenability to automation, and cost-effectiveness. Additionally, electrochemical techniques offer easy applicability and require minimum amounts and numbers of chemicals, and electrons to facilitate wastewater treatment instead of using chemicals and micro-organisms (Aji et al. 2012). The sludge generated by electrochemical processes is relatively low and tends to be readily settleable and easy to de-water because it is composed of mainly metallic oxides/hydroxides (Varank & Sabuncu 2015). Among the electrochemical technologies, electrocoagulation (EC) and electro-Fenton (EF) are of particular interest because of their high pollutant removal efficiencies.

EC processes a direct the current source between metal electrodes immersed in wastewater causing the dissolution of metal electrodes into wastewater. Wide ranges of coagulated species and metal hydroxides that destabilize and aggregate the suspended particles or precipitate and adsorb dissolved contaminants are formed by the dissolved metal ions at an appropriate pH. The main processes occurring during EC are: (i) migration to an oppositely charged electrode (electrophoresis) and aggregation due to charge neutralization; (ii) the cation or hydroxyl ion (OH) forms a precipitate with the pollutant; (iii) the metallic cation interacts with OH to form a hydroxide, which has high adsorption properties, thus bonding to the pollutant (bridge coagulation); (iv) the hydroxides form larger lattice-like structures and sweeps through the water (sweep coagulation); (v) oxidation of pollutants to less toxic species; and (vi) removal by electroflotation or sedimentation and adhesion to bubbles (Chen 2004; Canizares et al. 2005; Katal & Pahlavanzadeh 2011).

In the case of using Al electrodes, monomeric species such as Al(OH)+2, Al(OH)2+2, Al2(OH)2+4, Al(OH)−4 and polymeric species such as Al6(OH)15+3, Al7(OH)17+4, Al8(OH)20+4, Al13O4(OH)24+7, and Al13(OH)34+5 are formed during the EC process (Can et al. 2003). In the case of using Fe electrodes Fe(OH)+2, Fe(OH)2+, Fe2(OH)2+4, Fe(OH)4, Fe(H2O)2+, Fe(H2O)5OH+2, Fe(H2O)4(OH)2+, Fe(H2O)8(OH)2+4, Fe2(H2O)6(OH)4+2 are formed.

In electrochemical processes, the use of hydroxyl radicals in aqueous medium has been suggested to promote the oxidation of pollutants present in the effluent to be treated (Andreozzi et al. 1999). The hydroxyl radical is an extremely powerful oxidant. Huang et al. (1993) concluded that the methods based on hydrogen peroxide to promote the formation of hydroxyl radical are the most promising approaches (Roa-Morales et al. 2007). The oxidizing power of H2O2 can be strongly enhanced using the EF method in acidic medium. A typical EF reaction should involve three key reactions: (1) the generation of H2O2 from dissolved oxygen on the surface of the cathode, (2) the generation of hydroxyl radicals (·OH) between H2O2 and Fe2+, and (3) the degradation of organic substance by the ·OH.

Since the classical optimization technique of changing one variable at a time to study the effects of variables on the response is a time consuming and expensive method and it does not represent the effect of interactions between different factors for multivariable systems, response surface methodology (RSM), a statistical experimental design, is used for the modeling and analysis of problems in which several variables effect a response of interest, with the objective of optimizing this response by performing a minimum number of experiments. Central composite design (CCD) is the most common design under RSM. It provides sufficient data on the effects of variables and offers useful data about direct, pairwise interaction and curvilinear variable effects (Mohajeri et al. 2010a, 2010b). Recently, the RSM method has been used to determine optimum parameters in different processes (Amr et al. 2014; Roosta et al. 2014; Witek-Krowiak et al. 2014; Sing et al. 2017) and different types of wastewater (Can 2014; Ghanbari & Moradi 2015; Varank & Sabuncu 2015; Thirugnanasambandham et al. 2016; Varank et al. 2016; Guvenc et al. 2017).

The aim of this research was to determine the optimal conditions for COD, phenol and Ca+2 removal from paper mill industry wastewater by EC and EF processes through RSM. Since the parameters of COD, phenol and Ca+2 take place in common pollutants in paper industry wastewater, these parameters have been chosen as responses. In addition, at high pH levels, Ca+2 can react with HCO3 alkalinity and form Ca hardness. Detrimental effects that high levels of organics and calcium hardness may cause in the paper industry are reported (Kim et al. 2003). Ca precipitations, as well as concentrated organic matters in paper industry wastewater, can form pitches, resulting in operation problems such as wire or pipe clogging and drainage rate reduction (Kim et al. 2003). The experimental runs were designed in accordance with CCD. Experimental data and model predicted data were analyzed via analysis of variance (ANOVA). Additionally, the total costs of EC and EF processes, including chemical cost and operational cost, were also reported at optimum conditions.

Experimental setup and procedure

A laboratory-scale Plexiglass reactor with 9 cm diameter and 13 cm height was used in the study. Electrode sets (two anode and two cathode electrodes) comprised of four monopolar parallel iron plates (6 cm width × 11.5 cm height and 0.2 cm thickness), each having an effective area of 46.2 cm2. The electrodes were placed 1.5 cm apart. A valve was installed at the bottom of the reactor to withdraw the precipitated material through a sludge chamber. The experimental set-up is shown in Figure 1. For each test, 500 mL of wastewater sample was used.
Figure 1

Experimental set-up.

Figure 1

Experimental set-up.

Close modal

Electrolyte solution (1,000 mg/L NaOH) was used, since salinity of the wastewater samples was not sufficient. The pH of the solution was adjusted to the desired value using 6 N H2SO4 and NaOH. All the experimental studies were conducted at room temperature and under atmospheric pressure. Before each run, electrodes were washed with acetone and the impurities on the aluminum electrode surfaces were removed by dipping in a solution freshly prepared by mixing 100 cm3 of HCl solution (35%) and 200 cm3 of hexamethylenetetramine aqueous solution (2.80%) for 5 min (Gengec et al. 2012; Varank & Sabuncu 2015).

Paper mill industry wastewater analytical procedures

Real wastewater obtained from a paper mill industry located in Halkali, Istanbul was used in this study. The characterisation of the wastewater is provided in Table 1. Paper mill industry effluents were preserved and all analyses were carried out in accordance with the standard methods of APHA (2005). COD was measured by a closed reflux titrimetric method and phenol concentrations were measured by a direct photometric Method (5530 D) of standard methods of APHA. Analysis of H2O2 was carried out by the permanganometric method (Kurt et al. 2007). Residual H2O2 was also measured in the supernatant to evaluate possible interference with COD. The required amount of H2O2 was added to the wastewater sample in each run based on the H2O2/COD ratio.

Table 1

Characteristics of paper mill industry wastewater

ParameterInfluent values
pH 5.55 
Conductivity (mS/cm) 6.97 
COD (mg/L) 12,717 
Suspended solids (mg/L) 98.75 
Total solids (mg/L) 12,120 
Chloride (mg/L) 541.83 
Phenol (mg/L) 20.41 
Ca+2 (mg/L) 1,667.8 
ParameterInfluent values
pH 5.55 
Conductivity (mS/cm) 6.97 
COD (mg/L) 12,717 
Suspended solids (mg/L) 98.75 
Total solids (mg/L) 12,120 
Chloride (mg/L) 541.83 
Phenol (mg/L) 20.41 
Ca+2 (mg/L) 1,667.8 

Ca+2 concentrations of the influent and effluent wastewater samples were analyzed by using Inductively Coupled Plasma-Optical Emission Spectroscopy. In the EF process, the study was carried out at lower pH values because pH plays a more important role in this process, as compared with the EC process. pH controls the production of hydroxyl radicals and the concentration of ferrous ions in the solution (Mohajeri et al. 2010a, 2010b; Varank et al. 2016). Removal efficiencies of COD, phenol and Ca+2 were calculated using Equation (1):
formula
1
where C0 and C (mg/L) are the initial and final concentrations of the contaminants, respectively.

Experimental design and statistical analysis

RSM, an experimental design, uses mathematical and statistical techniques to determine the relative significance of various factors in the presence of complex interactions (Gengec et al. 2012; Karichappan et al. 2014; Varank et al. 2016). CCD, the most popular class of second-order designs of RSM, was used as an experimental design in this study to optimize COD, phenol and Ca+2 removal from paper mill industry wastewater via EC and EF processes. In the present study, a three-factorial and a three-level central composite experimental design leading to a total number of 15 experiments was employed for the EC process and a four-factorial and a three-level central composite experimental design leading to a total number of 30 experiments was employed for the EF process. Statgraphics Centurion XVI.I software program was used for design, mathematical modeling, and optimization. Initial pH (X1), current density (mA/cm2) (X2) and operation time (min) (X3) were studied as variables (independence factors) for the EC process whereas initial pH (X1), current density (mA/cm2) (X2), operation time (min) (X3) and H2O2/COD (X4) ratio were studied as variables for the EF process. Each independent variable in both processes was coded and their actual values and ranges are shown in Table 2. COD, phenol and Ca+2 removal efficiencies (%) (Y1, Y2 and Y3, respectively) were considered to be dependent factors (responses).

Table 2

Experimental range and levels of the independent variables

ProcessFactorsOriginal factor (X)Coded factors
− 10+ 1
EC pH X1 
Current density (mA/cm2X2 32 64 96 
Time (min) X3 10 20 30 
EF pH X1 
Current density (mA/cm2X2 32 64 96 
Time (min) X3 10 20 30 
H2O2/COD ratio X4 0.5 1.25 
ProcessFactorsOriginal factor (X)Coded factors
− 10+ 1
EC pH X1 
Current density (mA/cm2X2 32 64 96 
Time (min) X3 10 20 30 
EF pH X1 
Current density (mA/cm2X2 32 64 96 
Time (min) X3 10 20 30 
H2O2/COD ratio X4 0.5 1.25 

For statistical calculations, the selected independent variables were converted into dimensionless codified values to allow the comparison of factors of different natures with different units and to decrease the error in the polynomial fit according to Equation (2):
formula
2
where Xi is the dimensionless coded value of ith independent variable; Xr is the value of Xi at the center point; ΔXi is the step change value.
The quadratic equation model (Montgomery 2001; Ahmad et al. 2007) for predicting the optimal point can be expressed according to Equation (3):
formula
3
where Y is the response, are the independent factors, is the constant coefficient, , , are the coefficients of the liner, quadratic and second-order effect, k represents the number of factors and is the random error.

ANOVA was used for graphical analyses of the data to obtain the interaction between the process variables and the responses. The quality of the fit polynomial model was expressed by the coefficient of determination R2, and its statistical significance was checked by the Fisher F-test in the same program. Model terms were examined by P value (probability) with 95% confidence level. Three-dimensional plots and their respective contour plots were obtained based on the effects of the levels of three factors (pH, current density and time) for EC and four factors (pH, current density, time and H2O2/COD ratio) for EF. The simultaneous interaction of the factors on the responses was studied from three-dimensional plots.

Cost analysis

In the EC and EF processes, energy and electrode costs were the two significant components of operational cost. Thus, the cost of optimum EC and EF processes was calculated using Equation (4):
formula
4
where ENC is electrical energy consumption (kWh/m3), ELC is electrode consumption (kg/m3), a (€/kWh) is the price of electrical energy and b (€/kg) is the price of electrode material in Turkish markets.
The electrical energy consumption was calculated using Equation (5):
formula
5
where U is the applied voltage (V), i is the current intensity (A), tEC is the operation time (s or min), v is the volume of the treated wastewater (m3).
The amount of electrode dissolved was calculated theoretically using Faraday's law as shown in Equation (6):
formula
6
where Mw is the molecular weight of iron electrode (g/mol), z is the number of electrons transferred (three for iron), F is Faraday's constant (96,487 C/mol) and v is the volume of the treated wastewater. The cost of chemical consumption was ignored since chemical consumption is not high in electrochemical processes and chemical cost is not a major component of the total cost.

Statistical analysis and model fitting by CCD

The actual design of experiments and responses for COD, phenol and Ca+2 removal by EC and EF optimization are presented in Tables 3 and 4. It can be seen from Tables 3 and 4 that the proposed empirical model is suitable for predicting COD, phenol and Ca+2 removals for both processes, revealing a reasonably good agreement. The second-order (quadratic) polynomial response surface model was applied to fit the experimental results obtained by CCD. Based on the experimental design results, the regression equations with the coded variables obtained for describing the COD, phenol and Ca+2 removal from paper mill industry wastewater by EC and EF processes are presented in Equations (7)–(12):
formula
7
formula
8
formula
9
formula
10
formula
11
formula
12
where X1, X2, X3 and X4 (the coefficients with one factor) represent the effects of the linear main factor, while X1*X2, X1*X3, X1*X4, X2*X3 and X2*X4 (the coefficients with two factors) and X12, X22, X32 and X42 (the coefficients with second-order terms) represent the interaction between the two factors and the quadratic effects, respectively. The positive sign of coefficients indicates a synergistic effect, whereas the negative sign of coefficients indicates an antagonistic effect (Bajpai et al. 2012; Varank & Sabuncu 2015). On the basis of the coefficients in Equations (7)–(12) it can be seen that COD and phenol removal efficiencies increase but Ca+2 removal efficiency decreases with an increase in pH and current density. COD and Ca+2 removal increases with an increase in operation time but phenol removal decreases with increased operation time in EC process. In the EF process, COD and Ca+2 removal increases but phenol removal decreases with an increase in pH and H2O2/COD ratio. COD, phenol and Ca+2 removal efficiencies increase as the current density increases. Operation time has a negative effect on COD removal but a positive effect on phenol and Ca+2 removal.
Table 3

The actual design of experiments and responses for COD, phenol and Ca+2 removal by EC optimization

Run no.Independent variables
Y1 (%) (COD removal efficiency)
Y2 (%) (phenol removal efficiency)
Y3 (%) (Ca+2 removal efficiency)
X1X2X3ActualPredictedActualPredictedActualPredicted
32 10 18.08 17.98 61.68 62.52 13.42 12.36 
96 10 22.8 24.36 72.42 74.32 18.25 18.53 
32 30 25.22 23.65 79.91 78.01 16.24 15.95 
96 30 28.26 28.35 84.27 83.42 23.45 24.51 
32 20 21.08 21.84 34.3 34.11 21.76 21.81 
96 20 23.33 22.43 41.94 40.70 26.87 25.57 
32 20 21.75 22.65 72.26 73.49 17.56 18.85 
96 20 33.90 33.14 83.93 84.11 29.88 29.82 
64 10 21.95 21.28 37.00 36.33 20.50 21.50 
10 64 30 21.03 21.83 45.42 47.50 25.32 25.55 
11 64 10 23.56 22.75 78.70 76.61 21.66 21.42 
12 64 30 31.22 31.88 89.37 90.03 27.95 26.94 
13 64 20 27.60 28.50 75.95 74.18 18.25 19.08 
14 64 20 28.90 28.50 74.38 74.18 19.45 19.08 
15 64 20 29.00 28.50 72.22 74.18 19.56 19.08 
Run no.Independent variables
Y1 (%) (COD removal efficiency)
Y2 (%) (phenol removal efficiency)
Y3 (%) (Ca+2 removal efficiency)
X1X2X3ActualPredictedActualPredictedActualPredicted
32 10 18.08 17.98 61.68 62.52 13.42 12.36 
96 10 22.8 24.36 72.42 74.32 18.25 18.53 
32 30 25.22 23.65 79.91 78.01 16.24 15.95 
96 30 28.26 28.35 84.27 83.42 23.45 24.51 
32 20 21.08 21.84 34.3 34.11 21.76 21.81 
96 20 23.33 22.43 41.94 40.70 26.87 25.57 
32 20 21.75 22.65 72.26 73.49 17.56 18.85 
96 20 33.90 33.14 83.93 84.11 29.88 29.82 
64 10 21.95 21.28 37.00 36.33 20.50 21.50 
10 64 30 21.03 21.83 45.42 47.50 25.32 25.55 
11 64 10 23.56 22.75 78.70 76.61 21.66 21.42 
12 64 30 31.22 31.88 89.37 90.03 27.95 26.94 
13 64 20 27.60 28.50 75.95 74.18 18.25 19.08 
14 64 20 28.90 28.50 74.38 74.18 19.45 19.08 
15 64 20 29.00 28.50 72.22 74.18 19.56 19.08 

X1: pH; X2: current density – mA/cm2; X3: electrolysis time – min.

Table 4

The actual design of experiments and responses for COD, phenol and Ca+2 removal by EF optimization

Run no.Independent variables
Y4 (%) (COD removal efficiency)
Y5 (%) (phenol removal efficiency)
Y6 (%) (Ca+2 removal efficiency)
X1X2X3X4ActualPredictedActualPredictedActualPredicted
32 10 0.5 50.65 48.89 71.03 72.71 28.86 26.73 
32 10 61.23 56.62 92.67 90.87 46.59 41.65 
96 10 0.5 56.23 57.94 91.29 91.13 34.49 38.20 
96 10 70.30 70.43 96.20 97.28 59.45 61.13 
32 30 0.5 54.35 55.03 91.15 90.20 33.70 34.77 
32 30 58.89 61.27 97.26 99.75 46.42 49.71 
96 30 0.5 66.52 63.30 95.42 97.85 33.34 32.40 
96 30 74.88 74.30 97.80 95.40 58.49 55.37 
32 10 0.5 24.50 23.62 67.35 69.71 20.77 20.12 
10 32 10 29.93 33.3 94.76 92.21 28.38 32.60 
11 96 10 0.5 41.15 38.97 88.89 86.29 33.89 33.87 
12 96 10 55.60 53.46 95.88 96.79 59.22 54.38 
13 32 30 0.5 34.40 34.47 84.63 83.43 24.21 25.80 
14 32 30 45.89 42.72 97.21 97.32 45.80 38.32 
15 96 30 0.5 45.88 49.04 87.50 89.25 24.56 25.73 
16 96 30 60.09 62.05 92.94 91.14 40.86 46.26 
17 64 20 0.5 38.53 40.91 95.54 92.19 35.16 31.32 
18 64 20 48.72 51.28 98.29 102.21 43.30 49.04 
19 32 20 1.25 34.71 38.52 87.46 87.28 21.32 26.30 
20 96 20 1.25 51.58 52.71 92.65 93.40 39.08 36.00 
21 64 10 1.25 40.41 46.68 90.12 91.16 34.44 37.36 
22 64 30 1.25 55.37 54.04 97.55 97.08 38.34 37.32 
23 64 20 1.25 57.28 62.51 98.78 96.37 42.02 43.36 
24 64 20 1.25 44.04 43.75 89.75 92.73 34.93 35.50 
25 64 20 1.25 52.36 47.87 95.84 95.38 35.57 36.51 
26 64 20 1.25 51.18 47.87 96.67 95.38 38.54 36.51 
27 64 20 1.25 48.08 47.87 95.15 95.38 37.09 36.51 
28 64 20 1.25 49.36 47.87 95.45 95.38 36.06 36.51 
29 64 20 1.25 50.23 47.87 94.65 95.38 39.12 36.51 
30 64 20 1.25 50.89 47.87 96.26 95.38 38.45 36.51 
Run no.Independent variables
Y4 (%) (COD removal efficiency)
Y5 (%) (phenol removal efficiency)
Y6 (%) (Ca+2 removal efficiency)
X1X2X3X4ActualPredictedActualPredictedActualPredicted
32 10 0.5 50.65 48.89 71.03 72.71 28.86 26.73 
32 10 61.23 56.62 92.67 90.87 46.59 41.65 
96 10 0.5 56.23 57.94 91.29 91.13 34.49 38.20 
96 10 70.30 70.43 96.20 97.28 59.45 61.13 
32 30 0.5 54.35 55.03 91.15 90.20 33.70 34.77 
32 30 58.89 61.27 97.26 99.75 46.42 49.71 
96 30 0.5 66.52 63.30 95.42 97.85 33.34 32.40 
96 30 74.88 74.30 97.80 95.40 58.49 55.37 
32 10 0.5 24.50 23.62 67.35 69.71 20.77 20.12 
10 32 10 29.93 33.3 94.76 92.21 28.38 32.60 
11 96 10 0.5 41.15 38.97 88.89 86.29 33.89 33.87 
12 96 10 55.60 53.46 95.88 96.79 59.22 54.38 
13 32 30 0.5 34.40 34.47 84.63 83.43 24.21 25.80 
14 32 30 45.89 42.72 97.21 97.32 45.80 38.32 
15 96 30 0.5 45.88 49.04 87.50 89.25 24.56 25.73 
16 96 30 60.09 62.05 92.94 91.14 40.86 46.26 
17 64 20 0.5 38.53 40.91 95.54 92.19 35.16 31.32 
18 64 20 48.72 51.28 98.29 102.21 43.30 49.04 
19 32 20 1.25 34.71 38.52 87.46 87.28 21.32 26.30 
20 96 20 1.25 51.58 52.71 92.65 93.40 39.08 36.00 
21 64 10 1.25 40.41 46.68 90.12 91.16 34.44 37.36 
22 64 30 1.25 55.37 54.04 97.55 97.08 38.34 37.32 
23 64 20 1.25 57.28 62.51 98.78 96.37 42.02 43.36 
24 64 20 1.25 44.04 43.75 89.75 92.73 34.93 35.50 
25 64 20 1.25 52.36 47.87 95.84 95.38 35.57 36.51 
26 64 20 1.25 51.18 47.87 96.67 95.38 38.54 36.51 
27 64 20 1.25 48.08 47.87 95.15 95.38 37.09 36.51 
28 64 20 1.25 49.36 47.87 95.45 95.38 36.06 36.51 
29 64 20 1.25 50.23 47.87 94.65 95.38 39.12 36.51 
30 64 20 1.25 50.89 47.87 96.26 95.38 38.45 36.51 

X1: pH; X2: current density – mA/cm2; X3: electrolysis time – min; X4: H2O2/COD.

The significance of regression parameters was evaluated using ANOVA. The ANOVA analysis indicated that all three variables for EC and four variables for EF and their interactions had a significant effect on the models. The quadratic model statistical results are summarized in Table 5. The results show a high reliability in the estimation of COD, phenol and Ca+2 removal efficiencies (R2 values of 0.959, 0.993 and 0.969 for EC process and 0.936, 0.934 and 0.890 for EF process on COD, phenol and Ca+2 removal, respectively, were obtained). High R2 values ensure satisfactory adjustment of the quadratic model to the experimental data. It can be inferred that only 4.1% (COD removal), 0.7% (phenol removal) and 3.1% (Ca+2 removal) of the variability in the responses were not explained by the models in EC process, whereas 6.4% (COD removal), 6.6% (phenol removal) and 11% (Ca+2 removal) of variability in the responses were not explained by the models in the EF process. The model F-values imply that the models are significant for COD, phenol and Ca+2 removal efficiencies (Table 5). The associated P value is used to estimate whether statistical significance is large enough (Amani-Ghadima et al. 2013; Sridhar et al. 2016). The lower values of Prob > F (<0.05) imply that the model terms are significant, whereas higher values than 0.1 show that the model terms are insignificant at 95% probability level (Muhamad et al. 2013; Kim 2016). As can be seen from Table 5, response surface quadratic models for our parameters were significant at the 5% confidence level since P values were less than 0.05.

Table 5

The ANOVA of regression parameters of the predicted response surface quadratic model

ProcessModelR2Adjusted R2Sum of squaresMean squareF-valueProb > F
EC COD 0.959 0.887 263.59 29.28872 13.23956 0.0054750 
Phenol 0.993 0.983 4,670.9 518.9889 91.43967 0.0000519 
Ca+2 0.969 0.915 288.40 32.04528 17.84515 0.0027340 
EF COD 0.936 0.877 3,464.8 247.4846 15.83704 0.0000016 
Phenol 0.934 0.872 1415.1 101.0793 15.17116 0.0000022 
Ca+2 0.890 0.788 2,513.1 179.5108 8.725383 0.0000771 
ProcessModelR2Adjusted R2Sum of squaresMean squareF-valueProb > F
EC COD 0.959 0.887 263.59 29.28872 13.23956 0.0054750 
Phenol 0.993 0.983 4,670.9 518.9889 91.43967 0.0000519 
Ca+2 0.969 0.915 288.40 32.04528 17.84515 0.0027340 
EF COD 0.936 0.877 3,464.8 247.4846 15.83704 0.0000016 
Phenol 0.934 0.872 1415.1 101.0793 15.17116 0.0000022 
Ca+2 0.890 0.788 2,513.1 179.5108 8.725383 0.0000771 

ANOVA results for EC and EF processes are given in Tables 6 and 7, respectively. For EC process, F values were found to be 13.24, 91.44 and 17.85 for COD, phenol and Ca+2 removal, respectively, and P values were determined to be lower than 0.05 for all parameters. It can be concluded that the model for COD and Ca+2 removals are statistically significant, while the model for phenol removal is highly significant, explaining the relationship between the independent variables and the responses. It can be seen from Table 6 that all three variables have a significant effect on COD removal, while the quadratic coefficients of current density and operation time have a significant effect on COD removal. Additionally, only interaction between current density and operation time has an insignificant effect on COD removal. pH has a highly significant effect on phenol removal, whereas current density and electrolysis time have significant effects. The only coefficient in the quadratic coefficients that has a highly significant effect on phenol removal is pH. Current density and operation time have insignificant effects on phenol removal. It can be concluded from Table 6 that all interacting coefficients have an insignificant effect on phenol removal. The only linear and quadratic coefficient that has a significant effect on Ca+2 removal is determined to be pH. All interacting coefficients except interaction between pH and current density have insignificant effects on Ca+2 removal (Table 6).

Table 6

ANOVA results of the quadratic models for EC process

SourceSum of squaresDfMean squareF-valueProb > FRemark
COD removal (%) 
 Model 263.590 29.28872 13.23956 0.0054750 
 X1 66.3552 66.3552 29.99 0.0028 
 X2 61.3832 61.3832 27.74 0.0033 
 X3 46.7545 46.7545 21.13 0.0059 
 X1 X1 6.40913 6.40913 2.900 0.1495 NS 
 X1 X2 24.5025 24.5025 11.07 0.0208 
 X1 X3 18.4041 18.4041 8.320 0.0344 
 X2 X2 17.3467 17.3467 7.840 0.0380 
 X2 X3 0.70560 0.70560 0.320 0.5967 NS 
 X3 X3 27.7710 27.7710 12.55 0.0165 
Total error 11.0629 2.21257    
Total (corr.) 274.647 14     
R2 = 0.959 
Phenol removal (%) 
 Model 4,670.9 518.9889 91.43967 0.0000519 HS 
 X1 3,427.92 3,427.92 603.75 <0.0001 HS 
 X2 148.006 148.006 26.07 0.0038 
 X3 302.211 302.211 53.23 0.0008 
 X1 X1 724.899 724.899 127.67 <0.0001 HS 
 X1 X2 4.06023 4.06023 0.72 0.4363 NS 
 X1 X3 1.26563 1.26563 0.22 0.6567 NS 
 X2 X2 15.7321 15.7321 2.77 0.1569 NS 
 X2 X3 10.1761 10.1761 1.79 0.2383 NS 
 X3 X3 22.1782 22.1782 3.91 0.1051 NS 
Total error 28.3887 5.67775    
Total (corr.) 4,698.90 14     
R2 = 0.993 
Ca+2 removal (%) 
 Model 288.40 32.04528 17.84515 0.0027340 
 X1 0.845 0.845 0.47 0.5233 NS 
 X2 108.56 108.56 60.44 0.0006 
 X3 45.7446 45.7446 25.47 0.0039 
 X1 X1 110.646 110.646 61.60 0.0005 
 X1 X2 12.996 12.996 7.24 0.0433 
 X1 X3 0.540225 0.540225 0.30 0.6070 NS 
 X2 X2 1.09001 1.09001 0.61 0.4712 NS 
 X2 X3 1.4161 1.4161 0.79 0.4153 NS 
 X3 X3 1.8265 1.8265 1.02 0.3595 NS 
Total error 8.98114 1.79623    
Total (corr.) 297.343 14     
R2 = 0.969 
SourceSum of squaresDfMean squareF-valueProb > FRemark
COD removal (%) 
 Model 263.590 29.28872 13.23956 0.0054750 
 X1 66.3552 66.3552 29.99 0.0028 
 X2 61.3832 61.3832 27.74 0.0033 
 X3 46.7545 46.7545 21.13 0.0059 
 X1 X1 6.40913 6.40913 2.900 0.1495 NS 
 X1 X2 24.5025 24.5025 11.07 0.0208 
 X1 X3 18.4041 18.4041 8.320 0.0344 
 X2 X2 17.3467 17.3467 7.840 0.0380 
 X2 X3 0.70560 0.70560 0.320 0.5967 NS 
 X3 X3 27.7710 27.7710 12.55 0.0165 
Total error 11.0629 2.21257    
Total (corr.) 274.647 14     
R2 = 0.959 
Phenol removal (%) 
 Model 4,670.9 518.9889 91.43967 0.0000519 HS 
 X1 3,427.92 3,427.92 603.75 <0.0001 HS 
 X2 148.006 148.006 26.07 0.0038 
 X3 302.211 302.211 53.23 0.0008 
 X1 X1 724.899 724.899 127.67 <0.0001 HS 
 X1 X2 4.06023 4.06023 0.72 0.4363 NS 
 X1 X3 1.26563 1.26563 0.22 0.6567 NS 
 X2 X2 15.7321 15.7321 2.77 0.1569 NS 
 X2 X3 10.1761 10.1761 1.79 0.2383 NS 
 X3 X3 22.1782 22.1782 3.91 0.1051 NS 
Total error 28.3887 5.67775    
Total (corr.) 4,698.90 14     
R2 = 0.993 
Ca+2 removal (%) 
 Model 288.40 32.04528 17.84515 0.0027340 
 X1 0.845 0.845 0.47 0.5233 NS 
 X2 108.56 108.56 60.44 0.0006 
 X3 45.7446 45.7446 25.47 0.0039 
 X1 X1 110.646 110.646 61.60 0.0005 
 X1 X2 12.996 12.996 7.24 0.0433 
 X1 X3 0.540225 0.540225 0.30 0.6070 NS 
 X2 X2 1.09001 1.09001 0.61 0.4712 NS 
 X2 X3 1.4161 1.4161 0.79 0.4153 NS 
 X3 X3 1.8265 1.8265 1.02 0.3595 NS 
Total error 8.98114 1.79623    
Total (corr.) 297.343 14     
R2 = 0.969 

HS: highly significant; S: significant; NS: not significant.

X1: pH; X2: current density – mA/cm2; X3: electrolysis time – min.

Table 7

ANOVA results of the quadratic models for EF process

SourceSum of squaresDfMean squareF-valueProb > FRemark
COD removal (%) 
 Model 3,464.8 14 247.4846 15.83704 0.0000016 HS 
 X1 1,583.91 1,583.91 101.33 <0.0001 HS 
 X2 905.677 905.677 57.94 <0.0001 HS 
 X3 243.984 243.984 15.61 0.0013 
 X4 483.812 483.812 30.95 <0.0001 HS 
 X1 X1 71.6938 71.6938 4.59 0.0490 
 X1 X2 39.6585 39.6585 2.54 0.1320 NS 
 X1 X3 22.2077 22.2077 1.42 0.2518 NS 
 X1 X4 4.03006 4.03006 0.26 0.6190 NS 
 X2 X2 13.1707 13.1707 0.84 0.3732 NS 
 X2 X3 0.61231 0.612306 0.04 0.8458 NS 
 X2 X4 22.6814 22.6814 1.45 0.2470 NS 
 X3 X3 16.0684 16.0684 1.03 0.3267 NS 
 X3 X4 2.19781 2.19781 0.14 0.7129 NS 
 X4 X4 8.15976 8.15976 0.52 0.4811 NS 
Total error 234.456 15 15.6304    
Total (corr.) 3,699.45 29     
R2 = 0.936 
Phenol removal (%) 
 Model 1,415.1 14 101.0793 15.17116 0.0000022 HS 
 X1 59.3687 59.3687 8.91 0.0093 
 X2 168.361 168.361 25.27 0.0002 
 X3 157.65 157.65 23.66 0.0002 
 X4 452.102 452.102 67.85 <0.0001 HS 
 X1 X1 1.76644 1.76644 0.27 0.6142 NS 
 X1 X2 3.36722 3.36722 0.51 0.4881 NS 
 X1 X3 14.1376 14.1376 2.12 0.1658 NS 
 X1 X4 18.879 18.879 2.83 0.1130 NS 
 X2 X2 65.701 65.701 9.86 0.0067 
 X2 X3 115.778 115.778 17.37 0.0008 
 X2 X4 144.12 144.12 21.63 0.0003 
 X3 X3 4.08531 4.08531 0.61 0.4458 NS 
 X3 X4 74.1321 74.1321 11.12 0.0045 
 X4 X4 8.62271 8.62271 1.29 0.2732 NS 
Total error 99.9549 15 6.66366    
Total (corr.) 1,515.41 29     
R2 = 0.934 
Ca+2 removal (%) 
 Model 2,513.1 14 179.5108 8.725383 0.0000771 HS 
 X1 278.008 278.008 13.51 0.0022 
 X2 423.696 423.696 20.59 0.0004 
 X3 0.00760556 0.00760556 0.00 0.9849 NS 
 X4 1,413.88 1,413.88 68.70 <0.0001 HS 
 X1 X1 22.0116 22.0116 1.07 0.3174 NS 
 X1 X2 5.25556 5.25556 0.26 0.6207 NS 
 X1 X3 5.51076 5.51076 0.27 0.6124 NS 
 X1 X4 5.91706 5.91706 0.29 0.5997 NS 
 X2 X2 74.4431 74.4431 3.62 0.0766 NS 
 X2 X3 191.338 191.338 9.30 0.0081 
 X2 X4 64.3605 64.3605 3.13 0.0973 NS 
 X3 X3 1.78375 1.78375 0.09 0.7725 NS 
 X3 X4 0.00105625 0.00105625 0.00 0.9944 NS 
 X4 X4 34.8917 34.8917 1.70 0.2125 NS 
Total error 308.693 15 20.5796    
Total (corr.) 2,821.37 29     
R2 = 0.890 
SourceSum of squaresDfMean squareF-valueProb > FRemark
COD removal (%) 
 Model 3,464.8 14 247.4846 15.83704 0.0000016 HS 
 X1 1,583.91 1,583.91 101.33 <0.0001 HS 
 X2 905.677 905.677 57.94 <0.0001 HS 
 X3 243.984 243.984 15.61 0.0013 
 X4 483.812 483.812 30.95 <0.0001 HS 
 X1 X1 71.6938 71.6938 4.59 0.0490 
 X1 X2 39.6585 39.6585 2.54 0.1320 NS 
 X1 X3 22.2077 22.2077 1.42 0.2518 NS 
 X1 X4 4.03006 4.03006 0.26 0.6190 NS 
 X2 X2 13.1707 13.1707 0.84 0.3732 NS 
 X2 X3 0.61231 0.612306 0.04 0.8458 NS 
 X2 X4 22.6814 22.6814 1.45 0.2470 NS 
 X3 X3 16.0684 16.0684 1.03 0.3267 NS 
 X3 X4 2.19781 2.19781 0.14 0.7129 NS 
 X4 X4 8.15976 8.15976 0.52 0.4811 NS 
Total error 234.456 15 15.6304    
Total (corr.) 3,699.45 29     
R2 = 0.936 
Phenol removal (%) 
 Model 1,415.1 14 101.0793 15.17116 0.0000022 HS 
 X1 59.3687 59.3687 8.91 0.0093 
 X2 168.361 168.361 25.27 0.0002 
 X3 157.65 157.65 23.66 0.0002 
 X4 452.102 452.102 67.85 <0.0001 HS 
 X1 X1 1.76644 1.76644 0.27 0.6142 NS 
 X1 X2 3.36722 3.36722 0.51 0.4881 NS 
 X1 X3 14.1376 14.1376 2.12 0.1658 NS 
 X1 X4 18.879 18.879 2.83 0.1130 NS 
 X2 X2 65.701 65.701 9.86 0.0067 
 X2 X3 115.778 115.778 17.37 0.0008 
 X2 X4 144.12 144.12 21.63 0.0003 
 X3 X3 4.08531 4.08531 0.61 0.4458 NS 
 X3 X4 74.1321 74.1321 11.12 0.0045 
 X4 X4 8.62271 8.62271 1.29 0.2732 NS 
Total error 99.9549 15 6.66366    
Total (corr.) 1,515.41 29     
R2 = 0.934 
Ca+2 removal (%) 
 Model 2,513.1 14 179.5108 8.725383 0.0000771 HS 
 X1 278.008 278.008 13.51 0.0022 
 X2 423.696 423.696 20.59 0.0004 
 X3 0.00760556 0.00760556 0.00 0.9849 NS 
 X4 1,413.88 1,413.88 68.70 <0.0001 HS 
 X1 X1 22.0116 22.0116 1.07 0.3174 NS 
 X1 X2 5.25556 5.25556 0.26 0.6207 NS 
 X1 X3 5.51076 5.51076 0.27 0.6124 NS 
 X1 X4 5.91706 5.91706 0.29 0.5997 NS 
 X2 X2 74.4431 74.4431 3.62 0.0766 NS 
 X2 X3 191.338 191.338 9.30 0.0081 
 X2 X4 64.3605 64.3605 3.13 0.0973 NS 
 X3 X3 1.78375 1.78375 0.09 0.7725 NS 
 X3 X4 0.00105625 0.00105625 0.00 0.9944 NS 
 X4 X4 34.8917 34.8917 1.70 0.2125 NS 
Total error 308.693 15 20.5796    
Total (corr.) 2,821.37 29     
R2 = 0.890 

HS: highly significant; S: significant; NS: not significant.

X1: pH; X2: current density – mA/cm2; X3: electrolysis time – min; X4: H2O2/COD.

According to ANOVA results of the quadratic models for the EF process shown in Table 7, the F and P values were determined to be 15.84 and 0.0000016 for COD removal, 15.17 and 0.0000022 for phenol removal, 8.73 and 0.0000771 for Ca+2 removal, respectively, implying that the models are highly significant. The pH, current density and operation time have a highly significant effect whereas H2O2/COD ratio has a significant effect on COD and phenol removal. All quadratic coefficients except pH and all interactive coefficients have insignificant effects on COD removal. It can be seen from Table 7 that the only quadratic coefficient that has a significant effect on phenol removal is current density. Additionally, interactions between current density and operation time, current density and H2O2/COD ratio, operation time and H2O2/COD ratio have a significant effect on phenol removal. According to the ANOVA results, H2O2/COD ratio has a highly significant effect, whereas pH and current density in linear coefficients and interaction between current density and operation time in interactive coefficients has a significant effect on Ca+2 removal.

For the graphical explanation of the interactions, three dimensional plots of the regression models are presented in Figures 25. The response surface plots obtained from the software provide a three-dimensional view of the COD, phenol and Ca+2 removals surface with different combinations of independent variables. All response surface plots have clear peaks, meaning that the optimum conditions for maximum values of the responses are attributed to all variables in the design space. Three-dimensional surface plots indicate that, at optimum operating conditions, COD, phenol and Ca+2 removals were found to be maximum but moving away from these points shows a reduction in removal efficiencies, meaning that neither an increase or decrease in any of the tested variables is desired. As can be seen in Figures 2 and 3, the COD, phenol and Ca+2 removal efficiencies were determined to be much higher by the EF process than the EC process.
Figure 2

Three-dimensional response surface graphs for EC treatment of paper mill industry wastewater.

Figure 2

Three-dimensional response surface graphs for EC treatment of paper mill industry wastewater.

Close modal
Figure 3

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (COD removal).

Figure 3

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (COD removal).

Close modal
Figure 4

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (phenol removal).

Figure 4

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (phenol removal).

Close modal
Figure 5

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (Ca+2 removal).

Figure 5

Three-dimensional response surface graphs for EF treatment of paper mill industry wastewater (Ca+2 removal).

Close modal

Based on response surface and desirability functions, optimum conditions for maximum COD, phenol and Ca+2 removal from paper mill industry wastewater were determined. Optimized conditions for COD, phenol and Ca+2 removals are given in Table 8. An additional experiment was carried out at optimum conditions in order to confirm the accuracy of the predicted models and the reliability of the optimum combination. The experimental values were found to agree well with the predicted ones. Under these conditions, 34.7% of COD, 92.32% of phenol and 32.36% Ca+2 removals for EC and 74.31% of COD, 99.99% of phenol and 61.18% Ca+2 removals were obtained.

Table 8

Optimum operating conditions for EC and EF processes

FactorEC
EF
CODPhenolCa+2CODPhenolCa+2
pH 9.00 8.57 9.00 2.00 2.13 2.00 
Current density (mA/cm296.00 91.15 95.27 96.00 58.30 93.12 
Time (min) 27.55 30.00 30.00 30.00 28.36 10.00 
H2O2/COD – – – 2.00 2.00 2.00 
Predicted removal (%) 34.70 92.32 32.36 74.31 99.99 61.18 
Experimental removal (%) 33.50 91.20 33.85 75.60 97.25 60.50 
FactorEC
EF
CODPhenolCa+2CODPhenolCa+2
pH 9.00 8.57 9.00 2.00 2.13 2.00 
Current density (mA/cm296.00 91.15 95.27 96.00 58.30 93.12 
Time (min) 27.55 30.00 30.00 30.00 28.36 10.00 
H2O2/COD – – – 2.00 2.00 2.00 
Predicted removal (%) 34.70 92.32 32.36 74.31 99.99 61.18 
Experimental removal (%) 33.50 91.20 33.85 75.60 97.25 60.50 

Initial pH is an important operating factor influencing the performance of EC process (Chen et al. 2000; Adhoum & Monser 2004; Chen 2004). It has been reported that the highest removal rates are achievable (Koparal et al. 2008; Tchamango et al. 2010) at higher pH values, especially over neutral values. Bensadok et al. (2008) determined maximum removal rates at neutral pH 6–7, which is in agreement with many previous studies (Kobya et al. 2003; Sanchez-Calvo et al. 2003; Inan et al. 2004).

In the EF process, the study was carried out at lower pH values (in the range of pH values 2–4). Since pH controls the production of hydroxyl radicals and the concentration of ferrous ions in the solution, pH plays a more important role as compared with the EC process (Mohajeri et al. 2010a, 2010b; Varank et al. 2016). As pH increases, the iron ions, especially the Fe+3 ions, precipitate inhibiting the regeneration of ferrous ions and the amount of hydroxyl radicals generated decrease. Moreover, hydrogen peroxide is unstable under alkaline conditions and rapidly decomposes to water and oxygen as pH increases. Stable hydroxyl radicals, that have high oxidizing potential, are produced at pH values of 2–4 (Zhang et al. 2007; Mohajeri et al. 2010a, 2010b; Varank et al. 2016). On the other hand, H2O2 cannot be decomposed to OH by Fe2+, the rate of reaction between H2O2 and Fe2+ decreases and the removal rate decreases at pH values lower than 2.

The electrical current causes the dissolution of metal electrodes into wastewater in EC and EF processes. The dissolved metal ions, at an appropriate pH, can form wide ranges of coagulated species and metal hydroxides that destabilize and aggregate the suspended particles or precipitate and adsorb dissolved contaminants (Bazrafshan et al. 2013). The current supplied to the electrochemical reactor is usually expressed in terms of current density (Kim et al. 2003; Babu et al. 2007). Current density is expected to exhibit a strong effect on the EC process, especially on COD removal: the higher the current density, the shorter the treatment and the higher removal efficiency. At higher cell current values, the amount of metal oxidized increases, resulting in a greater amount of hydroxide flocs for the removal of pollutants. It is well known that electrical potential not only determines the coagulant dosage rate but also the bubble production rate and size and the flocs growth which can influence the treatment efficiency of the EC process (Bazrafshan et al. 2013). As the cell current increases, the bubble density increases and the size of the bubbles decreases, resulting in a faster removal of pollutants (Kim et al. 2003).

As a result, it can be said that optimum performances for COD, phenol and Ca+2 removal were achieved at about pH 9 for EC and at pH 2 for EF. In this study it can be seen from Table 8 that the highest removal rates were obtained at high pH values for the EC process and at low pH values for the EF process but at high current density values for both processes.

The total cost of EC and EF processes applied to paper mill industry wastewater for COD, phenol and Ca+2 removal at optimum conditions are given in Table 9. Operational cost and chemical cost values of the two processes were taken into account in calculation of the total cost. Total cost of COD removal at optimum conditions was determined to be 6.62 and 9.04 (€/m3), that of phenol removal was determined to be 8.74 and 6.01 (€/m3) whereas the total cost of Ca+2 removal was found to be 9.15 and 4.26 (€/m3) for the EC and EF processes, respectively. It can be concluded that no significant difference was observed in the total cost of the two processes.

Table 9

Total cost of COD, phenol and Ca+2 removal at optimum conditions in EC and EF processes

FactorEC
EF
CODPhenolCa+2CODPhenolCa+2
Operational cost (€/m36.44 6.65 6.95 7.01 4.01 2.23 
Chemical cost (€/m32.2 2.09 2.2 0.42 0.39 0.42 
H2O2 cost (€/m3– – – 1.61 1.61 1.61 
FactorEC
EF
CODPhenolCa+2CODPhenolCa+2
Operational cost (€/m36.44 6.65 6.95 7.01 4.01 2.23 
Chemical cost (€/m32.2 2.09 2.2 0.42 0.39 0.42 
H2O2 cost (€/m3– – – 1.61 1.61 1.61 

In this study, CCD and RSM were adopted to model and optimize the performance of EC and EF processes and to determine the optimal experimental conditions for COD, phenol and Ca+2 removal from paper mill industry wastewater. The results of the quadratic model developed in this study showed a good agreement between experimental and predicted values. ANOVA showed good coefficient of determination values. At optimum conditions, 34.7% of COD, 92.32% of phenol and 32.36% Ca+2 removal was achieved by the EC process, whereas 74.31% of COD, 99.99% of phenol and 61.18% Ca+2 removal was achieved by the EF process. The EF process offered higher removal efficiencies for COD, phenol and Ca+2 but no significant difference was observed between the total costs of EC and EF processes. Hence, the EF process can be recommended as a powerful technique for paper mill wastewater treatment.

Aghdam
M. A.
Kariminia
H.-R.
Safari
S.
2016
Removal of lignin, COD, and color from pulp and paper wastewater using electrocoagulation
.
Desal. Water Treat.
57
,
9698
9704
.
American Public Health Association, APHA
2005
Standard Methods for the Examination of Water and Wastewater, 21st edn. American Water Works Association and Water Environment Federation, Washington, DC, USA
.
Andreozzi
R.
Caprio
V.
Insola
A.
Marrotta
R.
1999
Advanced oxidation processes (AOP) for water purification and recovery
.
Catal. Today
53
,
51
59
.
Asaithambi
P.
Aziz
A. R. A.
Sajjadi
B.
Daud
W. M. A. B. W.
2016
Sono assisted electrocoagulation process for the removal of pollutant from pulp and paper industry effluent
.
Environ. Sci. Pollut. Res.
24
,
5168
5178
.
Babu
R. R.
Bhadrinarayana
N. S.
Begum
K. M. M. S.
Anantharaman
N.
2007
Treatment of tannery wastewater by electrocoagulation
.
J. Uni. Chem. Technol. Metall.
42
(
2
),
201
206
.
Bazrafshan
E.
Moein
H.
Mostafapour
F. K.
Nakhaie
S.
2013
Application of electrocoagulation process for dairy wastewater treatment
.
J. Chem.
2013
,
1
8
.
Bensadok
K.
Benammara
S.
Lapicque
F.
Nezzal
G.
2008
Electrocoagulation of cutting oil emulsions using aluminium plate electrodes
.
J. Hazard. Mater.
152
,
423
430
.
Canizares
P.
Carmona
M.
Lobato
J.
Martinez
F.
Rodrigo
M. A.
2005
Electrodissolution of aluminum electrodes in electrocoagulation processes
.
Ind. Eng. Chem. Res.
44
,
4178
4185
.
Chen
X.
Chen
G.
Yue
P. L.
2000
Separation of pollutants from restaurant wastewater by electrocoagulation
.
Separ. Purif. Technol.
19
,
65
76
.
Demirel
B.
Yenigun
O.
Onay
T. T.
2005
Anaerobic treatment of dairy wastewaters: a review
.
Process Biochem.
40
,
2583
2595
.
Habets
L. H.
Knelissen
J. H.
1985
Application of the UASB reactor for anaerobic treatment of paper and board mill effluent
.
Water Sci. Technol.
17
,
61
75
.
Haghighat
F.
2015
Wastewater treatment in the pulp-and-paper industry: A review of treatment processes and the associated greenhouse gas emission
.
J. Environ. Manage
158
,
146
157
.
Inan
H.
Dimoglo
A.
Simsek
H.
Karpuzcu
M.
2004
Olive oil mill wastewater treatment by means of electrocoagulation
.
Separ. Purif. Technol.
36
,
23
31
.
Karichappan
T.
Venkatachalam
S.
Jeganathan
P. M.
2014
Optimization of electrocoagulation process to treat grey wastewater in batch mode using response surface methodology
.
J. Environ. Health Sci. Eng.
12
(
29
),
1
8
.
Koparal
A. S.
Yildiz
Y.
Keskinler
B.
Demircioglu
N.
2008
Effect of initial pH on the removal of humic substances from wastewater by electrocoagulation
.
Separ. Purif. Technol.
59
,
175
182
.
Kushwaha
J. P.
Srivastava
V. C.
Mall
I. D.
2010
Organics removal from dairy wastewater by electrochemical treatment and residue disposal
.
Separ. Purif. Technol.
76
,
198
205
.
Matt
D. Z.
Habet
L. H. A.
1987
The upflow anaerobic sludge blanket wastewater treatment system: a technological review
.
Pulp Paper Can.
88
(
11
),
60
64
.
Mohajeri
S.
Aziz
H. A.
Isa
M. H.
Zahed
M. A.
Adlan
M. N.
2010b
Statistical optimization of process parameters for landfill leachate treatment using electro-Fenton technique
.
J. Hazard. Mater.
176
,
749
758
.
Montgomery
D. C.
2001
Design and Analysis of Experiments
, 5th edn.
John Wiley and Sons
,
New York, NY, USA
.
Pokhrel
D.
Viraraghavan
T.
2004
Treatment of pulp and paper mill wastewater – a review
.
Sci. Total Environ.
333
,
37
58
.
Rajeshwar
K.
Ibanez
J.
Swain
G. M.
1994
Electrochemistry and the environment
.
J. Appl. Electrochem.
24
,
1077
1091
.
Roa-Morales
G.
Campos-Medina
E.
Aguilera-Cotero
J.
Bilyeu
B.
Barrera-Diaz
C.
2007
Aluminum electrocoagulation with peroxide applied to wastewater from pasta and cookie processing
.
Separ. Purif. Technol.
54
,
124
129
.
Sanchez-Calvo
L.
Leclerc
J. P.
Tanguy
G.
Cames
M. C.
Paternotte
G.
Valentin
G.
Rostan
A.
Lapicque
F.
2003
An electrocoagulation unit for the purification of soluble oil wastes of high COD
.
Environ. Progress Sustain. Energy
22
,
57
65
.
Sridhar
R.
Sivakumar
V.
Thirugnanasambandham
K.
2016
Response surface modeling and optimization of upflow anaerobic sludge blanket reactor process parameters for the treatment of bagasse based pulp and paper industry wastewater
.
Desal. Water Treat.
57
,
4345
4356
.
Tchamango
S.
Nanseu-Njiki
C. P.
Ngameni
E.
Hadjiev
D.
Darchen
A.
2010
Treatment of dairy effluents by electrocoagulation using aluminium electrodes
.
Sci. Total Environ.
408
,
947
952
.
Tchobanoglous
G.
Burton
F. L.
Stensel
H. D.
2003
Wastewater Engineering Treatment and Reuse
.
Metcalf and Eddy, McGraw-Hill
,
New York, NY, USA
.
Thomson
G.
Swain
J.
Kay
M.
Forster
C. F.
2001
The treatment of pulp and paper mill effluent: a review
.
Bioresour. Technol.
77
(
3
),
275
286
.
Van Langerak
E. P. A.
Hamelers
H. V. M.
Lettinga
G.
1997
Influent calcium removal by crystallization reusing anaerobic effluent alkalinity
.
Water Sci. Technol.
36
,
341
348
.
Zhang
H.
Fei
C.
Zhang
D.
Tang
F.
2007
Degradation of 4-nitrophenol in aqueous medium by electro-Fenton method
.
J. Hazard. Mater.
145
,
227
232
.