Application of Response Surface Methodology for Optimization of Treatment for an Aged Landfill Leachate Using Fenton's Oxidation Reagent

Abstract

Landfill leachates can contain organic pollutants that may have an adverse impact on water bodies receiving the leachates unless some form of treatment reduces the pollutants to acceptable levels (often measured as biochemical oxygen demand and chemical oxygen demand [COD]). Age of the landfill can dictate treatment methods with younger landfill leachates that are often being treated using cheaper biotreatment processes and older landfills often requiring chemical and/or physical treatment due to the biorecalcitrance of the organics in these older landfill leachates. Fenton's reagent oxidation is a process with potential to significantly reduce COD from landfill leachates. Fenton's reagent oxidation reactions were carried out at room temperature (20 ± 5°C) and atmospheric pressure. Experiments were conducted by adding the necessary amount of FeSO₄•7H₂O and 30% (w/w) H₂O₂ to the leachate. Samples of reaction mixture were collected at specified times for COD analysis; 6 N sodium hydroxide solution was added to samples to bring sample pH above 7.0; and MnO₂ was added to quench the reaction. Response surface central composite design was applied to optimize the Fenton's process condition for max COD reduction. Time, temperature, pH, and mole ratio of H₂O₂ to Fe²⁺ play an important role in the COD reduction. Optimum values of time, temperature, pH, and mole ratio of H₂O₂ to Fe²⁺ were found to be 4 h and 15 min, 60°C, 2.8, and 143 for the maximum of 91% COD reduction. Experimental values were in good agreement with response surface model predicted values.

Introduction

Sanitary landfills are the primary method that is currently used for disposal of municipal solid waste in many countries (Deng and Englehardt, 2006). However, an inevitable consequence of landfills is the generation of leachates (Shabiimam and Dikshit, 2012), which is water generated over time, rainwater enters into the landfill, percolates through the wastes within the landfill cell, and becomes contaminated with the original wastes and the degradation by-products of these waste via the biochemical processes that are active in landfills (Umar et al., 2010). Leachates are often collected within the landfill system via collection reservoirs (or ponded onto the liner) and pumped outside of the landfill into holding ponds, then treated for discharge (Salem et al., 2008).

Many factors, including waste type, waste compaction, landfill design, regional climate, and site hydrology (Umar et al., 2010), affect the quality and quantity of leachates. However, the quality of the leachates is strongly influenced by the age of landfills.

Generally, the leachates are classified based on the landfill age as young/active (0–5 years), or mid-aged (5–15 years), or old/matured (15 years and above). Young/active landfills can produce leachates with chemical oxygen demand (COD) levels in excess of 10,000 mg/L. The young to mid-aged landfills generally produce leachates with COD levels in the 6,000 to 10,000 mg/L range (Renou et al., 2008; Foo and Hameed, 2009). Old/matured landfills produce leachates with reduced COD levels that taper down with age as in situ degradation processes decompose the wastes (∼3,000 at 20 years).

In any case, collected leachates are often treated before discharge, in particular, leachates containing high levels of COD to protect the environment (Žgajnar Gotvajn et al., 2011).

Several components of the leachates are often refractory and toxic (Mahmud et al., 2012). Organic pollutants within leachates are often quantified by using totalized oxygen-use equivalence methods, such as biochemical oxygen demand (BOD) and chemical oxygen demand (COD) and/or measured total organic carbon concentrations. These commonly used pollution potential analytical methods are considered key assessment tools for assessing the environmental threat of releasing the landfill leachates along with their concentration reduction used as acceptable effluent goals during treatment (Shabiimam and Dikshit, 2012).

Selection of the best treatment method depends primarily on the composition and flow rate of the leachates (Lin et al., 2011). Biological treatment is the most frequently used method, mainly due to its simplicity, cost-effectiveness, and high efficiency in removal of biodegradable matter (Umar et al., 2010). It is effective in treating leachates from new to mid-aged landfills but often inadequate in treating leachates derived from “mature” facilities (Barbusiński and Pieczykolan, 2010). Recently, biological and physical-chemical treatments have been used together successfully in the treatment of matured leachates (Wiszniowski et al., 2006; Kamaruddin et al., 2015). However, the leachates containing biologically recalcitrant compounds, with the ratio of BOD₅ to COD less than 0.5, are not efficiently treated with biological processes (Ha et al., 2008).

For such leachates, chemical processes such as advanced oxidation processes (AOPs) have been proposed (Oller et al., 2011). Among the potential candidate AOP that may be employed, Fenton's reagent oxidation is widely used to improve leachates' biodegradability and treatability due to its simple operation and low costs (Yilmaz et al., 2010; Shabiimam and Dikshit, 2012). The factors that affect the performance of Fenton's reagent oxidation are reaction time, pH, mole ratio of H₂O₂ to Fe²⁺, Fe²⁺ dosage, and temperature (Zhang et al., 2009; Mohajeri et al., 2010).

The conventional experimental method of studying a process does not depict the combined effect of all the factors involved; it is time consuming and requires a large number of experiments to determine optimum levels, which may or may not be reliable. These limitations can be eliminated by simultaneously varying all the parameters by using statistical designed experiments such as the response surface methodology—central composite design (RSM-CCD) (Ponnusamy and Subramaniam, 2013). RSM-CCD is a collection of statistical and mathematical techniques that are useful for developing, improving, and optimizing processes that can be used to evaluate the relative significance of all the factors involved in the process, even in the presence of complex interactions with reduced variations, time, and cost (Box and Wilson, 1951; Myers et al., 2016).

Even though several studies reported on the treatment of landfill leachates with high COD values ranging from 570 mg/L to 34,920 mg/L, there is a lack in the systematic evaluation of optimum operating conditions for Fenton's treatment (Li et al., 2010; Cortez et al., 2011; Aygun et al., 2012; Ghanbarzadeh Lak et al., 2012; Ahmadian et al., 2013; Badawy et al., 2013; Silva et al., 2015). Optimum conditions for Fenton's treatment strongly depend on the characteristics of the leachate (Ghanbarzadeh Lak et al., 2012); moreover, recalcitrant, mature leachates with low COD values have received little research attention. In this study, a leachate produced by a mature landfill cell aged over 30–35 years with a low COD concentration (344 mg/L) and BOD/COD ratio of 0.2 was used to optimize the Fenton's reaction process conditions by using response surface central composite factorial design methodology for maximum COD reduction efficiency.

The goal is to reduce the landfill leachate COD level to below 50 mg/L. The targeted COD value is below the landfill leachate discharge standard. The COD level of landfill leachate discharge standard varies from 90 to 200 mg/L (Lin and Chang, 2000; Kurniawan et al., 2006; Wang et al., 2012). The RSM-CCD was selected to determine the effect of experimental parameters and their interactions for achieving maximum COD reduction and, thus, optimize the Fenton's reagent process toward treating the leachate.

Materials and Methods

Materials

A leachate generated within a mature landfill cell (30–35 years) was collected from the St. Landry Parish Solid Waste Disposal District, Washington, LA 70589 and brought to the Environmental Engineering Laboratory at the University of Louisiana at Lafayette where it was stored in a refrigerator at 4°C until its use during testing. The results of the leachate characterization analyzed by St. Landry Parish Solid Waste Disposal District are given in Table 1. The results presented in Table 1 were the average value of the results collected for the period of 6 months. Chemicals used in this work were an aqueous hydrogen peroxide stock solution (30% H₂O₂, w/w), ferrous sulfate, sulfuric acid, and sodium hydroxide. The hydrogen peroxide stock and ferrous sulfate (0.1 N) solutions were obtained from Aqua Solutions, Deer Park, TX. The peroxide test strips (EMD Millipore MQuant™ Peroxide Test Strips from fisher scientific) were used to check the hydrogen peroxide concentrations. The sulfuric acid and sodium hydroxide were obtained from Fisher Scientific (www.fishersci.com).

Table 1.

Characteristics of Landfill Leachate

Parameter	Value (mg/L)
TSS	38.00
Ammonia nitrate	2.55
Chlorides	227.50
Sulfates	10.05
TDS	763.50
DO	8.16
Alpha Terpineol	0.004125
Benzoic acid	0.0103
p-Cresol	0.00415
Zinc	0.005
Phenol	0.0041

DO, dissolved oxygen; TDS, total dissolved solids; TSS, total suspended solids.

Response surface central composite design

Response surface central composite statistical experiment design was employed to investigate the effects of the four independent variables on the response functions. The reaction time (X₁), temperature (X₂), pH (X₃), and mole ratio of H₂O₂ to Fe²⁺ (X₄) were chosen as the independent input variables. The ranges of independent input variables were: reaction time, 1–6 h; pH, 1.0–10.0; temperature, 20–60°C; and mole ratio of H₂O₂ to Fe²⁺, 3.3–328.4. The variable ranges chosen were based on the preliminary studies conducted in our lab and literature information (Zhang et al., 2005; Aygun et al., 2012; Zazouli et al., 2012; Benradi et al., 2016).

Independent variables, experimental range, and levels for landfill leachate are given in Table 2. The low, midpoint, and high levels of each variable were designated as 1, 0, and +1, respectively. For statistical calculations, the variables X_i were coded as x_i according to the following relationship: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} { x_i } = { \frac { { X_i } - { X_o } } { { \delta _X } } } \tag { 1 } \end{align*} \end{document}

where x_i is the dimensionless coded value of the variable X_i, X_o is the value of the X_i at the center point, and δ_X is the step change.

Table 2.

Experimental Range and Levels of Independent Process Variables

	Range and level
Independent variable	−1	0	+1
Time (h) (X₁)	1	3.5	6
Temperature (°C) (X₂)	20	40	60
pH⁺ (X₃)	1.0	5.5	10.0
Mole ratio of H₂O₂ to Fe²⁺ (mg/mg) (X₄)	3.3	165.8	328.4

A 2⁴ full factorial experimental design with seven replicates at the center point, eight experiments at axial points, and thus a total of 31 experiments as illustrated in Table 3 were employed in this study. The alpha value was chosen as face centered (α = 1). When α = 1, the axial points are placed on the “cube” portion of the design. This is an appropriate choice when the cube points of the design are at the operational limits. The center point replicates were chosen to verify any change in the estimation procedure, as a measure of precision property.

Table 3.

Central Composite Design Matrix with Experimental and Predicted Values

Coded values				Uncoded values
Time (h)	Temp (°C)	pH	Mole ratio of H₂O₂ to Fe ²⁺	Time (h)	Temp (°C)	pH	Mole ratio of H₂O₂ to Fe ²⁺	% COD removal experimental	% COD removal predicted
0	1	0	0	3.5	60	5.5	166	78.5	79.1
−1	−1	−1	−1	1.0	20	1.0	3	67.2	62.9
0	0	0	0	3.5	40	5.5	166	81.1	77.9
1	1	−1	−1	6.0	60	1.0	3	78.5	77.9
−1	−1	1	1	1.0	20	10.0	328	29.4	30.1
0	0	0	0	3.5	40	5.5	166	81.1	77.9
1	−1	1	1	6.0	20	10.0	328	51.2	45.5
−1	1	1	−1	1.0	60	10.0	3	25.0	20.4
0	0	0	0	3.5	40	5.5	166	81.1	77.9
1	1	1	−1	6.0	60	10.0	3	25.0	23.8
0	0	0	0	3.5	40	5.5	166	81.1	77.9
0	0	0	−1	3.5	40	5.5	3	67.2	70.2
−1	−1	−1	1	1.0	20	1.0	328	61.0	58.2
0	0	0	0	3.5	40	5.5	166	74.7	77.9
1	−1	−1	1	6.0	20	1.0	328	70.9	73.6
0	0	1	0	3.5	40	10.0	166	34.6	40.0
−1	0	0	0	1.0	40	5.5	166	64.0	68.3
1	1	1	1	6.0	60	10.0	328	14.1	19.1
0	0	0	1	3.5	40	5.5	328	63.1	65.5
1	−1	−1	−1	6.0	20	1.0	3	75.0	78.3
0	0	0	0	3.5	40	5.5	166	81.1	77.9
−1	1	−1	1	1.0	60	1.0	328	68.6	69.9
1	0	0	0	6.0	40	5.5	166	76.7	77.7
1	1	−1	1	6.0	60	1.0	328	75.9	73.2
−1	−1	1	−1	1.0	20	10.0	3	31.7	34.8
0	0	−1	0	3.5	40	1.0	166	81.1	81.1
0	0	0	0	3.5	40	5.5	166	81.1	77.9
−1	1	1	1	1.0	60	10.0	328	16.6	15.7
0	−1	0	0	3.5	20	5.5	166	81.7	86.5
−1	1	−1	−1	1.0	60	1.0	3	71.5	74.6
1	−1	1	−1	6.0	20	10.0	3	52.0	50.2

R² = 97.8.

COD, chemical oxygen demand.

The percent COD reduction (Fenton's reagent oxidation effectiveness) achieved for landfill leachate test influent was taken as a dependent response variable. The behavior of the system is explained by the following quadratic equation: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} Y = \;{ \beta _0} + \sum { \beta _i}{x_i} + \sum { \beta _{ii}}x_i^2 + \sum { \beta _{ij}}{x_i}{x_j} \tag{2} \end{align*} \end{document}

where β₀ is the offset term, β_i is the linear effect, β_ii is the squared effect, β_ij is the interaction effect, and x_i is the dimensionless coded value of the variable X_i. The results of the experimental design are studied and interpreted by MINITAB 14 (statistical software to estimate the response of the dependent variable). The graphical flow chart for the application of response surface methodology for optimization of treatment conditions for an aged landfill leachate using Fenton's oxidation reagent is given in Fig. 1.

FIG. 1.

Graphical flow chart for application of response surface methodology for optimization.

Analytic methods and experimental procedure

Initial and final COD values of the samples were measured by the HACH instruments method (Ranges 3 to 150 mg/L, USEPA Reactor Digestion Method; Method 8000, originally from Standard Method 5220 D) by using a HACH COD Digester (DRB 200). COD standards were used to validate the HACH measurements. Samples were diluted with distilled and deionized water prepared in the Environmental Engineering Laboratory to bring down the COD levels of samples below 150 mg/L (since the method and standards are for the range 3–150 mg/L). The COD standard recovery rates were 150 ± 4 mg/L. The COD reduction was calculated as follows: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \% \;COD \;reduction = { \frac { { { \left[ { COD } \right] } _ { initial } } - { { \left[ { COD } \right] } _ { final } } } { { { \left[ { COD } \right] } _ { initial } } } } \;\times\;100 \tag { 3 } \end{align*} \end{document}

Initial BOD₅ of the sample was measured by using the standard method: Procedure 5210 B (via a YSI DO meter equipped with a stirred, membrane O₂ BOD bottle probe). The pH of the leachate was measured by using a XL 50 Dual Channel pH/Ion/Conductivity Meter (Fisher Scientific, Pittsburgh, PA).

Fenton's reagent oxidation reactions were carried out in 500 mL magnetically stirred batch reactors (400 mL working volume) at room temperature (20 ± 5°C) and atmospheric pressure. Response surface central composite design specified concentrations of Fe²⁺ and H₂O₂ in the leachate were achieved by first adding the necessary amount of FeSO₄•7H₂O to the leachate, followed by the addition of requisite amounts of 30% (w/w) H₂O₂. Samples of the reaction mixture were collected at specified times of the design; 6 N sodium hydroxide solution was added to the samples to bring sample pH above 7.0 (to ensure accuracy of COD measurements); and MnO₂ was added to quench the reaction. All the experiments were conducted in triplicate as per the design developed with RSM-CCD, and the results given in Table 3 are their average values.

Results and Discussion

Reaction mechanism

Initial pH, COD, and BOD values of the leachate were 8.3, 344 mg/L, and 69 mg/L, respectively. Thus, the BOD/COD ratio for the leachate was 0.2. The presence of trace amounts of Alpha Terpineol, Benzoic Acid, and p-Cresol (Table 1) in the landfill leachate would play a critical role in COD reduction as well as in determining the optimum conditions of Fenton oxidation. Treatment of leachate by the Fenton's reagent oxidation is the most effective method. Fenton's reagent oxidation reaction was first proposed in the 1890s by Henry Fenton. In the Fenton's Reagent Oxidation, hydrogen peroxide is added to wastewater in the presence of ferrous salt, generating species that are strongly oxidative with respect to the organic compounds present (Qiang et al., 2003). In this reaction, hydroxyl free radicals are generally deemed to be the key species (Wu et al., 2011). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} F{e^{2 + }} + {H_2}{O_2} \to F{e^{3 + }} + \cdot OH + O{H^ - } \;{k_1} = 50 \sim 70 \;{M^{ - 1}}{s^{ - 1}} \tag{4} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} F{e^{3 + }} + {H_2}{O_2} \to F{e^{2 + }} + H{O_2} \cdot + {H^ + } \; \;{k_2} = 0.01 \;{M^{ - 1}}{s^{ - 1}} \tag{5} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \cdot OH + {H_2}{O_2} \to H{O_2} \cdot + {H_2}O \quad \quad {k_3} = 3.3 \times {10^7}{M^{ - 1}}{s^{ - 1}} \tag{6} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} H{O_2} \cdot + F{e^{3 + }} \to F{e^{2 + }} + {O_2}{H^ + } \quad \quad {k_4} < 2 \times {10^3}{M^{ - 1}}{s^{ - 1}} \tag{7} \end{align*} \end{document}

(Qiang et al., 2003) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} F{e^{3 + }} + H{O_2} \cdot \to F{e^{2 + }} + {O_2} + {H^ + } \quad \quad {k_5} = 1.2 \times {10^6}{M^{ - 1}}{s^{ - 1}} \tag{8} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} 2H{O_2} \cdot \to {H_2}{O_2} + {O_2} \ \quad \quad {k_6} = 8.3 \times {10^5}{M^{ - 1}}{s^{ - 1}} \; \; \tag{9} \end{align*} \end{document}

(Qiang et al., 2003) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} RH + \cdot OH \to {H_2}O + R ^\prime \to Further \;oxidation \; \quad \quad {k_7} \; \tag{10} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} H{O_2} \cdot + F{e^{2 + }} \to F{e^{3 + }} + H{O_2}^ - \quad \quad {k_8} = 1.3 \times {10^6}{M^{ - 1}}{s^{ - 1}} \; \; \; \tag{11} \end{align*} \end{document}

(Wu et al., 2011) \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \cdot OH + F{e^{2 + }} \to F{e^{3 + }} + O{H^ - } \quad \quad {k_9} = 3.2 \times {10^8}{M^{ - 1}}{s^{ - 1}} \; \; \; \tag{12} \end{align*} \end{document}

(Wu et al., 2011)

Here, RH is the organic matter and it is the source of COD in the leachate.

Statistical modeling and analysis of process

Experimental results showing the coded and un-coded value of the variables together with the % COD reduction efficiency for landfill leachate are given in Table 3. The % COD reduction was calculated by using Equation 3.

The experimental results (Table 3) allow the development of a mathematical model for % COD reduction (Y) as a function of the time, temperature, initial pH, and mole ratio of H₂O₂ to Fe²⁺ through RSM. Apart from the linear effect of the parameter for the % COD reduction, the RSM also gives an insight into the quadratic and interaction effect of the parameters. The un-coded regression model consists of a constant term, four first-order effects (X₁, X₂, X₃, and X₄), four second-order effects (X₁², X₂², X₃², and X₄²), and six interaction effects (terms in X₁X₂, X₁X₃, X₁X₄, X₂X₃, X₂X₄, and X₃X₄).

Statistical significance of each term in the model is summarized in Table 4. These analyses are done by means of Fisher's F test and Student's t-test. The Student's t-test was used to determine the significance of the regression coefficients of the parameters. The p-values are used as a tool to check the significance of each of the interaction among the variables, which, in turn, may indicate the patterns of the interactions among the variables. In general, the larger the magnitude of T and smaller the value of p, the more significant is the corresponding coefficient term (Montgomery, 2013).

Table 4.

Estimated Regression Coefficients for % Chemical Oxygen Demand Removal

Term	Coef	SE Coef	T	p
Constant	54.6202	9.68402	5.64	0
Time (h)	9.4435	3.2373	2.917	0.01
Temp (°C)	−0.5198	0.55273	−0.941	0.361
pH	7.6933	1.61081	4.776	0
Mole ratio of H₂O₂ to Fe²⁺	0.1228	0.03884	3.161	0.006
Time (h) × time (h)	−0.7769	0.4306	−1.804	0.09
Temp (°C) × temp (°C)	0.0122	0.00673	1.819	0.088
pH × pH	−0.8571	0.1329	−6.449	0
Mole ratio of H₂O₂ to Fe²⁺ × mole ratio of H₂O₂ to Fe²⁺	−0.0004	0.0001	−3.739	0.002
Time (h) × temp (°C)	−0.06	0.02168	−2.768	0.014
Time (h) × pH	0.0422	0.09634	0.438	0.667
Time (h) × mole ratio of H₂O₂ to Fe²⁺	0.0002	0.00267	0.081	0.936
Temp (°C) × pH	−0.0722	0.01204	−5.997	0
Temp (°C) × mole ratio of H₂O₂ to Fe²⁺	−0.0002	0.00033	−0.657	0.521
pH × mole ratio of H₂O₂ to Fe²⁺	−0.0006	0.00148	−0.382	0.708

R-Sq = 97.9%, R-Sq_(adj) = 96.1%.

SE, standard error.

The regression coefficient and the T and p-values for all the linear, quadratic, and interaction effects of the parameters are given in Table 4. From the high T values and very small p-values, it is observed that the coefficients for the linear effect and quadratic effect of all the factors are highly significant, except the linear effect of temperature. The high p-values of 0.361, 0.667, 0.936, 0.521, and 0.708 for the linear effect of temperature, the interaction effects of time and pH, time and mole ratio of H₂O₂ to Fe²⁺, temperature and mole ratio of H₂O₂ to Fe²⁺, and pH and mole ratio of H₂O₂ to Fe²⁺ did not seem to be statistically significant, respectively. However, the temperature has a positive effect (slight increase) on the reaction rate between hydrogen peroxide and Fe²⁺ ion, which increases the rate of generation of hydroxyl radicals at lower temperatures (20°C to 40°C).

On the other hand, the increase of temperature has a negative effect on the decomposition of hydrogen peroxide into water and oxygen and has a positive effect on the generation of hydroxyl radicals. Continued increase of temperature (beyond 40°C) has a negative effect on the conversion of hydrogen peroxide into hydroxyl radicals (Zhang et al., 2005; Hermosilla et al., 2009; Aygun et al., 2012). The interaction effect of time and pH, and time and mole ratio of H₂O₂ to Fe²⁺ was reported by Li et al. (2010) as significant. The interaction effect of pH and mole ratio of H₂O₂ to Fe²⁺ on COD reduction was reported as significant by Ghanbarzadeh Lak et al. (2012) and Li et al. (2010). But the significance level was not discussed. Statistical analysis resulted that the interaction model terms, X₁X₃, X₁X₄, X₂X₄, and X₃X₄ were insignificant. Hence, those model terms were eliminated and the final un-coded regression model (reduced second-order polynomial) for the % COD reduction efficiency (Y) is given in equation 13. \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \begin{split} {\rm{Y \;}} = & { \rm{ \;}}55.646827 + 9.711605{X_1} - 0.556102{X_2} \\ & + 7.747486{X_3} + 0.111676{X_4} - 0.776896X_1^2 \\ & + 0.012236X_2^2{ \rm{ \;}} - 0.857066X_3^2 - 0.000381X_4^2 \\ & -0.060000{X_1}{X_2} - 0.072222{X_2}{X_3}\end{split} \tag{13} \end{align*} \end{document}

Significance of the model was tested by using the analysis of variance (ANOVA) method, which is a statistical technique that subdivides the total variation in a set of data into component parts that are associated with specific sources of variation for the purpose of testing hypothesis on the parameters of the model (Ponnusamy and Subramaniam, 2013). According to the ANOVA (Table 5), the associated values of p lower than 0.01 indicate that the model is statistically significant. The model developed to explain the relationship between the factors and the response has very good agreement with the experimental value.

Table 5.

Analysis of Variance for % Chemical Oxygen Demand Removal

Source	DF	Seq SS	Adj SS	Adj MS	F	p
Regression	10	14225.7	14225.67	1422.57	90.23	0
Linear	4	8346.3	2724.25	681.06	43.2	0
Square	4	5059.3	5059.32	1264.83	80.23	0
Interaction	2	820	820	410	26.01	0
Residual Error	20	315.3	315.31	15.77
Lack of Fit	14	280.2	280.2	20.01	3.42	0.069
Pure Error	6	35.1	35.11	5.85
Total	30	14541

R² = 97.8%, R²_(adj) = 96.7%.

DF, degrees of freedom; MS, mean square; SS, sum of square.

The regression coefficient R² gives the proportion of the total variation in the response variable explained or accounted for by the predictors (X's) included in the model (Ahmadi et al., 2005). With a regression coefficient (R²) value of 0.978 (Table 5) for COD reduction, indicates that the models can explain at least 97.8% of the COD reduction. The overall model predictability for this aged landfill leachate and the conditions studied seems satisfactory compared with other research efforts (Li et al., 2010; Wu et al., 2010; Mohajeri et al., 2011; Ghanbarzadeh Lak et al., 2012), where their reported R² values ranged from 0.9322 to 0.99 for the COD removal efficiency models used.

The Response Surface Model predicted values for % COD reduction efficiency (using Equation 13) are given in Table 3. A good agreement between the experimental data and the model predicted data was also obtained. The calculated R² value of 0.978 indicates that the second-order polynomial model (Equation 13) is highly significant and adequate (97.8%) to represent the actual relationship between the response (% COD reduction efficiency) and the process variables.

Optimization of experimental conditions

The response surface plots to estimate the % COD reduction efficiency over independent variables are shown in Fig. 2. Surface plots are used to explore the potential relationship between the response and the two independent variables, and to locate the optimum. The peaks and valleys of the surface plots show the maximum and minimum value of the % COD reduction efficiency for the relative effects of two independent variables by keeping the other two variables at the midpoint values.

FIG. 2.

Surface plot of % COD removal (a) % COD removal versus time (h) and temperature (°C) (Hold values: pH—5.5 and mole ratio of H₂O₂ to Fe²⁺—165.8) (b) % COD removal versus time (h) and pH (Hold values: temperature—40°C and mole ratio of H₂O₂ to Fe²⁺—165.8) (c) % COD removal versus time (h) and mole ratio of H₂O₂ to Fe²⁺ (Hold values: temperature—40°C and pH—5.5) (d) % COD removal versus temperature (°C) and pH (Hold values: time—3.5 (h) and mole ratio of H₂O₂ to Fe²⁺—165.8) (e) % COD removal versus temperature (°C) and mole ratio of H₂O₂ to Fe²⁺ (Hold values: time—3.5 (h) and pH—5.5) (f) % COD removal versus pH and mole ratio of H₂O₂ to Fe²⁺ (Hold values: time—3.5 (h) and temperature—40°C). COD, chemical oxygen demand.

Figure 2a shows the effect time and temperature on % COD reduction. The maximum % COD remaining reduction occurs around 6 h and around 20°C while keeping pH and mole ratio constant at the center point. Initially, the % COD reduction was low and then increasing with time and decreasing with temperature, because at a higher temperature the concentration of hydroxyl radicals will be reduced due to incomplete hydrogen peroxide decomposition into water and oxygen (Yoon et al., 1998; Aygun et al., 2012).

Figure 2b illustrates the pH and the reaction time effect on % COD reduction. A higher COD reduction was obtained at the time range of 3–6 h and pH range of 1–4. A similar pH effect can be seen from Fig. 2d and 2f. At this pH range, a higher concentration of hydroxyl radicals is formed by the reaction involving the organometallic complex where hydrogen peroxide is either regenerated or wasted by an increased reaction rate (Sedlak and Andren, 1991). After that, pH has a negative impact on % COD reduction. Optimal pH for Fenton's reagent oxidation for landfill leachate treatment typically ranged between 2 and 4.5 (Zhang et al., 2005; Deng and Englehardt, 2006).

The effect of time and mole ratio of H₂O₂/Fe²⁺ on % COD reduction is given in Fig. 2c. From Fig. 2c, it is clear that the maximum % COD reduction can be obtained with the H₂O₂/Fe²⁺ mole ratio of 150. A similar trend was observed from Fig. 2e and 2f. At low molar ratios, the reaction rate follows second order and the rate approaches zero order when the molar ratio increases. This means that only reactions 4 and 12 occur at low molar ratios, but at high molar ratios, reactions 5, 8, and 11 would occur, which can change the reaction mechanism independent of hydrogen peroxide (Zhang et al., 2005). This H₂O₂/Fe²⁺ molar ratio of 150 is very high when compared with the ratios of 1–10 reported by Ghanbarzadeh Lak et al. (2012), Li et al. (2010), and Zhang et al. (2009). This is because of the high recalcitrant nature and the high scavenging effect of the leachate.

The response surface optimum conditions of the process variables were obtained in un-coded units using MINITAB 14. The optimum conditions for the maximum % COD reduction were time, 4 h 15 min; temperature, 60°C; pH, 2.8; and mole ratio (H₂O₂/Fe²⁺), 143. The predicted response at the optimum conditions from the model was 89.25% COD reduction efficiency. To verify this result, experiments were conducted in triplicate at the optimum conditions, and the percentage COD reduction thus obtained was 90.8%, which exhibits close agreement with the model.

Results in Fig. 3 show that when reaction time approaches 4 h, % COD reduction reaches the maximum point. With a further increase in reaction time, % COD reduction decreases with time. A similar phenomenon was observed by Yoon et al. (1998). This is because of the production of inhibitors that hindered the oxidation reaction (Gulsen and Turan, 2004) and the interaction effects of pH and mole ratio of H₂O₂/Fe²⁺. By increasing the temperature, % COD reduction efficiency was enhanced and the optimum temperature appears to be 60°C. Figure 3 shows that the impact of pH is very significant for % COD reduction. An optimum pH of 2.8 was determined in this study. Deng and Englehardt (2006) reported an optimum pH of 3.0, which has closer agreement with the results of this study.

FIG. 3.

Optimization plot of response surface design.

An optimum mole ratio of 143 was obtained for maximum COD reduction based on the RSM model developed from the experimental values. This value differs from the optimum mole ratio (H₂O₂ to Fe²⁺ = 100) reported by Lin et al. (2011). This difference is mainly due to the characteristics of leachate. Compared with similar studies published related to this work, our work is able to achieve comparable performance with the aged leachate that has low COD levels. The COD of the treated landfill leachate was 31.6 mg/L, which is below the targeted value of 50 mg/L as well as below the level of landfill leachate discharge standard (Lin and Chang, 2000; Kurniawan et al., 2006; Wang et al., 2012).

Conclusions

This investigation was performed to optimize process parameters for maximum COD reduction from a municipal landfill leachate by using Fenton's reagent oxidation by applying response surface methodology. The obtained regression model was used to predict the optimal conditions for maximum COD reduction. The developed model is adequate to explain the process behavior and predict the response with 97.8% confidence. Initial pH and molar ratio had a high significant effect on % COD reduction.

This study revealed that the maximum amount of COD that could be removed by Fenton's reagent oxidation was about 91% of the initial value. Such a maximum reduction was achieved by using the conditions of time: 4 h and 15 min; temperature: 60°C; pH: 2.8; and mole ratio of H₂O₂ to Fe²⁺: 143. From the results, it is clear that Fenton's oxidation reagent is effective for the degradation of recalcitrant compounds from stabilized landfill leachate. At the optimized conditions through statistically designed experiments, it can generate the treated effluent with a COD limit lower than 50 mg/L.

Footnotes

Acknowledgments

The authors are grateful for financial support from CLECO, and St. Landry Parish Solid Waste Disposal District in Washington, Louisiana, USA for providing the landfill leachate and chemical analysis of the leachate. This work was also partially supported by the Energy Institute of Louisiana (University of Louisiana).

Author Disclosure Statement

No competing financial interests exist.

References

Ahmadi

, Vahabzadeh

, Bonakdarpour

, Mofarrah

, and Mehranian

(2005). Application of the central composite design and response surface methodology to the advanced treatment of olive oil processing wastewater using Fenton's peroxidation. J. Hazard. Mater., 123, 187.

Ahmadian

, Reshadat

, Yousefi

, Mirhossieni

S.H.

, Zare

M.R.

, Ghasemi

S.R.

, Rajabi Gilan

, Khamutian

, and Fatehizadeh

(2013). Municipal leachate treatment by Fenton process: Effect of some variable and kinetics. J. Environ. Public Health, 2013, 169682.

Aygun

, Yilmaz

, Nas

, and Berktay

(2012). Effect of temperature on fenton oxidation of young landfill leachate: Kinetic assessment and sludge properties. Global NEST J. 14, 487–495.

Badawy

M.I.

, El-Gohary

, Gad-Allah

T.A.

, and Ali

M.E.M.

(2013). Treatment of landfill leachate by Fenton process: Parametric and kinetic studies. Desalination Water Treat. 51, 7323–7330.

Barbusiński

, and Pieczykolan

(2010). COD removal from landfill leachate using Fenton oxidation and coagulation. Agric. Civil Eng. Environ., 4, 93.

Benradi

, Yahyaoui

A.E.

, Bouhlassa

, Nounah

, Khamar

, and Ghrissi

(2016). Effect of pH and time on the leachate treatment by coagulation. J. Mater. Environ. Sci., 7, 1001.

Box

G.E.P.

, and Wilson

K.B.

(1951). On the experimental attainment of optimum conditions. J. R. Stat. Soc. Ser. B (Methodol.)., 13, 1.

Cortez

, Teixeira

, Oliveira

, and Mota

(2011). Evaluation of Fenton and ozone-based advanced oxidation processes as mature landfill leachate pre-treatments. J. Environ. Manage., 92, 749.

Deng

, and Englehardt

J.D.

(2006). Treatment of landfill leachate by the Fenton process. Water Res. 40, 3683.

10.

Foo

K.Y.

, and Hameed

B.H.

(2009). An overview of landfill leachate treatment via activated carbon adsorption process. J. Hazard. Mater., 171, 54.

11.

Ghanbarzadeh Lak

, Sabour

M.R.

, Amiri

, and Rabbani

(2012). Application of quadratic regression model for Fenton treatment of municipal landfill leachate. Waste Manag. 32, 1895.

12.

Gulsen

, and Turan

(2004). Treatment of sanitary landfill leachate using a combined anaerobic fluidized bed reactor and fenton's oxidation. Environ. Eng. Sci., 21, 627.

13.

D.Y.

, Cho

S.H.

, Kim

Y.K.

, and Leung

S.W.

(2008). A new approach to characterize biodegradation of organics by molecular mass distribution in landfill leachate. Water Environ. Res., 80, 748.

14.

Hermosilla

, Cortijo

, and Huang

C.P.

(2009). Optimizing the treatment of landfill leachate by conventional Fenton and photo-Fenton processes. Sci. Total Environ., 407, 3473.

15.

Kamaruddin

M.A.

, Yusoff

M.S.

, Aziz

H.A.

, Hung

Y.-T.

(2015). Sustainable treatment of landfill leachate. Appl. Water Sci., 5, 113.

16.

Kurniawan

T.A.

, Lo

W.-H.

, and Chan

G.Y.S.

(2006). Degradation of recalcitrant compounds from stabilized landfill leachate using a combination of ozone-GAC adsorption treatment. J. Hazard. Mater., 137, 443.

17.

, Zhou

, Sun

, and Lv

(2010). Application of response surface methodology to the advanced treatment of biologically stabilized landfill leachate using Fenton's reagent. Waste Manag. 30, 2122.

18.

Lin

S.H.

, and Chang

C.C.

(2000). Treatment of landfill leachate by combined electro-Fenton oxidation and sequencing batch reactor method. Water Res. 34, 4243.

19.

Lin

Y.-H.

, Juan

M.-L.

, and Hsien

H.-J.

(2011). Effects of temperature and initial pH on biohydrogen production from food-processing wastewater using anaerobic mixed cultures. Biodegradation, 22, 551.

20.

Mahmud

, Hossain

M.D.

, and Shams

(2012). Different treatment strategies for highly polluted landfill leachate in developing countries. Waste Manag. 32, 2096.

21.

Mohajeri

, Aziz

H.A.

, Isa

M.H.

, Zahed

M.A.

, and Adlan

M.N.

(2010). Statistical optimization of process parameters for landfill leachate treatment using electro-Fenton technique. J. Hazard. Mater., 176, 749.

22.

Mohajeri

, Aziz

H.A.

, Zahed

M.A.

, Mohajeri

, Bashir

M.J.K.

, Aziz

S.Q.

, Adlan

M.N.

, and Isa

M.H.

(2011). Multiple responses analysis and modeling of Fenton process for treatment of high strength landfill leachate. Water Sci. Technol., 64, 1652.

23.

Montgomery

D.C.

(2013). Design and Analysis of Experiments, 8th Edition. Hoboken, NJ: John Wiley & Sons, Inc.

24.

Myers

R.H.

, Montgomery

D.C.

, and Anderson-Cook

C.M.

(2016). Response Surface Methodology: Process and Product Optimization Using Designed Experiments, 4th Edition. Hoboken, NJ: John Wiley & Sons, Inc.

25.

Oller

, Malato

, and Sánchez-Pérez

J.A.

(2011). Combination of Advanced Oxidation Processes and biological treatments for wastewater decontamination—A review. Sci. Total Environ., 409, 4141.

26.

Ponnusamy

S.K.

, and Subramaniam

(2013). Process optimization studies of Congo red dye adsorption onto cashew nut shell using response surface methodology. Int. J. Ind. Chem., 4, 1.

27.

Qiang

, Chang

J.-H.

, and Huang

C.-P.

(2003). Electrochemical regeneration of Fe2+ in Fenton oxidation processes. Water Res. 37, 1308.

28.

Renou

, Givaudan

J.G.

, Poulain

, Dirassouyan

, and Moulin

(2008). Landfill leachate treatment: Review and opportunity. J. Hazard. Mater., 150, 468.

29.

Salem

, Hamouri

, Djemaa

, and Allia

(2008). Evaluation of landfill leachate pollution and treatment. Desalination, 220, 108.

30.

Sedlak

D.L.

, and Andren

A.W.

(1991). Oxidation of chlorobenzene with Fenton's reagent. Environ. Sci. Technol., 25, 777.

31.

Shabiimam

M.A.

, and Dikshit

A.K.

(2012). Treatment of municipal landfill leachate by oxidants. Am. J. Environ. Eng., 2, 1.

32.

Silva

T.F.

, Ferreira

, Soares

P.A.

, Manenti

D.R.

, Fonseca

, Saraiva

, Boaventura

R.A.

, and Vilar

V.J.

(2015). Insights into solar photo-Fenton reaction parameters in the oxidation of a sanitary landfill leachate at lab-scale. J. Environ. Manage., 164, 32.

33.

Umar

, Aziz

H.A.

, and Yusoff

M.S.

(2010). Trends in the use of Fenton, electro-Fenton and photo-Fenton for the treatment of landfill leachate. Waste Manag. 30, 2113.

34.

Wang

, Pelkonen

, and Kaila

(2012). Optimization of landfill leachate management in the aftercare period. Waste Manag. Res., 30, 789.

35.

Wiszniowski

, Robert

, Surmacz-Gorska

, Miksch

, and Weber

J.V.

(2006). Landfill leachate treatment methods: A review. Environ. Chem. Lett., 4, 51.

36.

, Zhou

, Qin

, Ye

, and Zheng

(2010). Modeling physical and oxidative removal properties of Fenton process for treatment of landfill leachate using response surface methodology (RSM). J. Hazard. Mater., 180, 456.

37.

, Zhou

, Zheng

, Ye

, and Qin

(2011). Mathematical model analysis of Fenton oxidation of landfill leachate. Waste Manag. 31, 468.

38.

Yilmaz

, Aygün

, Berktay

, and Nas

(2010). Removal of COD and colour from young municipal landfill leachate by Fenton process. Environ. Technol., 31, 1635.

39.

Yoon

, Cho

, and Kim

(1998). The characteristics of coagulation of fenton reaction in the removal of landfill leachate organics. Water Sci. Technol., 38, 209.

40.

Zazouli

M.A.

, Yousefi

, Eslami

, and Ardebilian

M.B.

(2012). Municipal solid waste landfill leachate treatment by fenton, photo-fenton and fenton-like processes: Effect of some variables. Iranian J. Environ. Health Sci. Eng., 9, 1.

41.

Žgajnar Gotvajn

, Zagorc-Končan

, and Cotman

(2011). Fenton's oxidative treatment of municipal landfill leachate as an alternative to biological process. Desalination, 275, 269.

42.

Zhang

, Choi

H.J.

, Canazo

, and Huang

C.P.

(2009). Multivariate approach to the Fenton process for the treatment of landfill leachate. J. Hazard. Mater., 161, 1306.

43.

Zhang

, Choi

H.J.

, and Huang

C.-P.

(2005). Optimization of Fenton process for the treatment of landfill leachate. J. Hazard. Mater., 125, 166.