Journal of Environmental Management 146 (2014) 9e15
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Subcritical water treatment of landfill leachate: Application of response surface methodology P. Kirmizakis a, C. Tsamoutsoglou b, B. Kayan c, D. Kalderis a, * a
Department of Environmental and Natural Resources Engineering, School of Applied Sciences, Technological and Educational Institute of Crete, Chania, Crete 73100, Greece b DEDISA S.A., Chania, Crete 73100, Greece c Department of Chemistry, Arts and Sciences Faculty, Aksaray University, Aksaray, Turkey
a r t i c l e i n f o
a b s t r a c t
Article history: Received 4 October 2013 Received in revised form 9 April 2014 Accepted 21 April 2014 Available online
Context: Leachate is the liquid formed when waste breaks down in the landfill and water filters through that waste. This liquid is highly toxic and can pollute the land, ground water and water ways. It is mandatory for landfills to protect against leachate in most countries worldwide. Controlling the pollutant loading, means reducing its quantity by containing or treating the waste to comply with certain discharge characteristics which are compatible with the receptor medium. Objective: This paper describes the reduction of the organic load of a mature landfill leachate using a novel experimental set-up that employs hydrogen peroxide under subcritical conditions and aims to establish this method as an effective alternative to currently used options. Response surface methodology was applied to optimize the treatment process and determine which of the following there parameters e temperature, residence time and hydrogen peroxide concentration e played the most important role. Method: The method employed is based on the use of laboratory-scale, stainless steel reactors, filled with the leachate and appropriate quantities of hydrogen peroxide. Under subcritical conditions (temperature in the range of 100e374 C and enough pressure to maintain the liquid state of water), hydrogen peroxide produces hydroxyl radicals which are highly reactive and oxidize the organic molecules of the leachate. Results: The highest COD decrease of 85% was experimentally observed at 300 C, 500 mM H2O2 and 180 min residence time. It was determined that the combination of oxidant concentration and temperature is the rate-determining factor, whereas residence time has a lesser effect on the process. Conclusions: A simple, quick, effective and environmentally-friendly method for the treatment of the organic load of landfill leachate was developed and optimized at laboratory scale. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Landfill leachate Subcritical water Superheated water Response surface methodology Central composite design Hydrogen peroxide
1. Introduction Landfill leachate can be defined as the liquid produced from the decomposition of solid waste and infiltration of rainwater in a landfill. It contains heavy metals, salts, nitrogen compounds and various types of organic substances (Schiopu and Gavrilescu, 2010). Generation of leachate occurs when moisture enters the solid waste in a landfill, dissolves the contaminants into liquid phase and
* Corresponding author. Tel.: þ30 2821023017; fax: þ30 2821023003. E-mail addresses: [email protected], [email protected] (D. Kalderis). http://dx.doi.org/10.1016/j.jenvman.2014.04.037 0301-4797/© 2014 Elsevier Ltd. All rights reserved.
produces moisture content sufficient to initiate liquid flow. This leachate is a high-polluting strength wastewater that has a major impact and influence on landfill design and its operation. Leachate composition varies from one landfill to another, and over space and time in a particular landfill with fluctuations that depend on short and long-term periods due to variations in climate, hydrogeology and waste composition (Keenan et al., 1984). Generally, leachate possesses high concentrations of ammonia and organic contaminants (measured in terms of chemical oxygen demand COD and biochemical oxygen demand BOD), halogenated hydrocarbons and heavy metals. In addition, leachate usually contains high concentrations of inorganic salts (mainly sodium chloride, carbonate and sulphate) and is dependent on the
10
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
composition of landfilled waste. The main environmental problems at landfill sites are the infiltration of leachate and its subsequent contamination of the surrounding land and aquifers. Improvements in landfill engineering are aimed at reducing leachate production, collection and treatment prior to discharge (Farquhar, 1989). The risks to human health and the environmental impact of landfill leachate contamination of ground- and surface water have been well documented (Slack et al., 2005; Macklin et al., 2011; Kwasniewska et al., 2012; Toufexi et al., 2013). An excess of 200 organic compounds have been identified in municipal landfill leachate, with more than 35 of them having the potential to cause harm to the environment and human health (Slack et al., 2005). Leachate needs to be pre-treated on site to meet the standards for its discharge into the sewer or its direct disposal into surface water. Leachate management is a complex task due to the highly variable nature of waste landfilled, type and design of the landfill site, its age, and climatic and seasonal variations in different regions. Hence, it is difficult to recommend treatment options merely based on leachate age but it is often necessary to consider each case individually. Treatment systems in recent years are sophisticated, reliable and able to consistently treat leachate to keep in line with specific discharge standards. Generally, high organic and ammonia loads are the key factors in leachate treatment. The feasibility of treating leachate to COD levels lower than 1000 mg/L is uncertain, since at these values COD is primarily composed of humic and fulvic acids. The effect of these substances on aquatic life of receiving waters is dependent extensively on specific cases. While developing a treatment sequence for leachate to be discharged into a river, Robinson et al. (2002) reported that COD levels of about 500 mg/L consisting of humic and fulvic acids did not adversely affect aquatic life. Worldwide, the issue of leachate treatment has been thoroughly investigated, but a universal solution has not yet been established. Therefore, there is a need to develop reliable and sustainable options to manage leachate generation and treatment effectively. While designing a treatment system, the process train must include techniques or unit operations to treat leachate produced from the landfill over a longer period. The most widely used treatment processes and their combinations have been thoroughly reviewed (Kjeldsen et al., 2002; Renou et al., 2008; Abbas et al., 2009; Foo and Hameed, 2009; Schiopu and Gavrilescu, 2010; Li et al., 2010; Oller et al., 2011; Ahmed et al., 2012). Subcritical water is hot water (>100 C) under enough pressure to maintain the liquid state. It is an environmentally friendly and inexpensive solvent that exhibits a wide range of properties that render it very effective in solvating and decomposing moderately polar or non-polar organic substances from a wide range of environmental matrices. Several studies have shown that subcritical water can decompose naturally-occurring substances and materials, such as complex amino acids, proteins and carbohydrates (sucrose, fructose, sorbose) and brown coal to produce more valuable and useful products. Additionally, subcritical water has been proven to decompose hazardous organic substances and materials such as residual reactive dyes (Daskalaki et al., 2011), fluorochemicals (Hori et al., 2008), explosives (Hawthorne et al., 2000; Kalderis et al., 2008), pentachlorophenol (Benedictus, 2007), dioxins (Hashimoto et al., 2004) and polyvinyl chloride (Takeshita et al., 2004). The scope of this work is to reduce the organic load of a mature landfill leachate using a novel experimental set-up that employs hydrogen peroxide under subcritical conditions and therefore establish this method as an effective alternative to currently used options. Using subcritical water, with or without oxidants is an effective way of degrading various environmental pollutants. Such
subcritical water set-up has not been used for the treatment of landfill leachate before. Compared to the studies mentioned earlier, this time the process aims at a highly complex wastewater consisting of a large number of organic and inorganic constituents of diverse nature. Hydrogen peroxide is used as an environmentallyfriendly oxidant, as it leaves no residues after treatment. In contrast to the dynamic set-ups that require considerable quantities of water and more complicated apparatus, the experiments described here are under static conditions i.e. no flow is required and no additional use of water. Additionally, monitoring of the process is not essential, since the oven can be pre-set at the required temperature and residence time. Finally, no pumping system is needed to maintain the system pressure, since pressure is automatically controlled by the steam/water equilibrium inside the reactor cell. When applying oxidative methods to decrease the organic load of the leachate, temperature, oxidant concentration and experimental time are the most important factors. Therefore, optimization of the influencing parameters is vital towards designing an effective degradation system. As a result, advanced statistical design has been widely employed for process characterization, optimization and modeling (Alaton et al., 2009; Singh et al., 2010). Experimental design methodologies have become a significant tool that have enabled us to gain a better understanding of a process in terms of the interactions among the parameters that need to be optimized. Response surface methodology (RSM) e a collection of mathematical and statistical techniques e has been found to be a useful aid for studying the effect of several factors that influence the response of a system as well as optimizing the variables of a wastewater treatment process. Furthermore, it is essential to choose an appropriate experimental design method that will evaluate the effects of the major parameters involved in the treatment method and their probable interactions, through the minimum number of experiments. RSM provides a large amount of information and is a more economical approach because a small number of experiments are performed for monitoring the interaction of the independent variables on the response. In conventional optimization, the increase in the number of experiments necessary to carry out the research, leads to an increase in time and expenses as well as an increase in the utilization of reagents and materials for experiments. Based on the above, the objectives of this study are the following: effectively reduce the organic load e as shown in COD measurements e of a mature landfill leachate using hydrogen peroxide in subcritical conditions apply a 5-level central composite experimental design (CCD) combined with RSM to optimize the various parameters and obtain maximum response, validate the suggested model and determine the optimum operational conditions, based on COD measurements
2. Materials and methods 2.1. Materials Landfill leachate was obtained from the Chania (Crete, Greece) municipal waste landfilling site with the help of the staff of DEDISA A.E. The sample was followed by an analytical certificate showing the levels of organic and inorganic constituents and other quality parameters (Table 1). The addition of the appropriate quantities of hydrogen peroxide (30% solution, SigmaeAldrich) was carried out
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
experiments were performed at the temperatures and residence times indicated by the central composite design (Table 2).
Table 1 Qualitative parameters of the landfill leachate sample obtained from DEDISA S.A. and other leachates described in literature. Parameter
Value
pH Conductivity (mS/cm1) Dissolved oxygen (mg L1) Total suspended solids (TSS, mg L1) NeNO3 (mg L1) NeNO2 (mg L1) 1 NeNHþ 4 (mg L ) P-PO4 (mg L1) TP (mg L1) BOD (mg L1) COD (mg L1) TOC (mg L1) 1 HCO 3 (mg L ) 1 SO2 4 (mg L ) Cl (mg L1) Naþ (mg L1) 1
This work
Alkassasbeh et al., 2009
8.1 22.5 3.4 447.50 16.76 0.50 2786.5 20.3 24.8 1434.4 12,360 698.3 10,676.7 8.5 2079.1 1644
7.4 26.3 0.78 nr1 nr nr 956 nr nr 1228 7600 nr nr nr nr nr
11
2.2. Methods 2.2.1. Central composite design and landfill leachate degradation optimization The design, mathematical modeling and optimization of this study were performed using Design Expert 8.0.5 software. Central composite design was used to model the RSM in this design. It is the most widely used experimental design for fitting a second-order response surface (Montgomery, 2009). It was first reported by Box and Wilson in 1951 and is well suited for fitting a quadratic surface, which usually works well for the process optimization € (Ozer et al., 2009). The independent variables (factors) used in this experimental study were temperature, H2O2 concentration and experimental time and are coded as x1, x2, and x3 respectively (Table 3). The CCD was conducted with 20 experiments, including 8 star points, 6 axial points corresponding to the alpha value and 6 replicates at the center points (Ghafari et al., 2009; Adlan et al., 2011). In the present study, such design was employed for determining the optimum conditions for landfill leachate organics degradation in subcritical water. The experimental range and levels of independent variables for landfill leachate degradation are given in Table 3. The experimental results were analyzed using Design Expert 8.0.5 and the regression model was proposed. In the optimization process, the responses can be simply related to the chosen factors by linear or quadratic models. A quadratic model, which also includes the linear model, is given below as Eq. (1).
7.6 14.7 2.52 nr nr nr 470 nr nr 866 3733 nr nr nr nr nr
not reported.
just before the experiment in order to avoid H2O2 photodecomposition. The experimental set-up and procedure is described in detail in Kalderis et al., 2008 and Daskalaki et al., 2011. Briefly, one type of small laboratory reactor was used for the degradation studies. The 25 ml reactors were constructed from (6-inches long, 0.64 inches i.d.) 316 stainless steel pipe with male national pipe threads (npt) and female end caps sealed with Teflon tape (Swagelok Company, USA). Each reactor was loaded with 20 ml landfill leachate, the appropriate quantity of H2O2 was added (according to the central composite design experiments e Table 2) and capped. This procedure left ca. 5 ml of headspace in the cell. All static (non-flowing) reaction cells must contain a sufficient headspace so that the pressure inside the cell is controlled by the steam/liquid equilibrium. A full cell must never be used since the pressures could reach several thousand bar. Cells were placed in a (pre-heated at the required temperature) GC oven (HewlettePackard 5890, series II) for heating. All
% Y ¼ bo þ
X
bi xi þ
X
bii xi2 þ
X
bij xi xj þ ε
(1)
where, bo is the constant coefficient, bi, bii and bij are the regression coefficient and xi, xj indicate the independent variables. ε represents the random error. The quality of the polynomial model was expressed by the coefficient of determination, namely, R2 and R2adj. The statistical significance was verified with adequate precision ratio and by the F-test (Rauf et al., 2008). According to the obtained experimental
Table 2 Central composite design experiments and experimental results. Experiment number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Experimental design
Experimental plan
T ( C)
H2O2(mM)
t (min)
x1
x2
x3
1 þ1 1 þ1 1 þ1 1 þ1 1.682 þ1.682 0 0 0 0 0 0 0 0 0 0
1 1 þ1 þ1 1 1 þ1 þ1 0 0 1.682 þ1.682 0 0 0 0 0 0 0 0
1 1 1 1 þ1 þ1 þ1 þ1 0 0 0 0 1.682 þ1.682 0 0 0 0 0 0
100 300 100 300 100 300 100 300 31.82 368.18 200 200 200 200 200 200 200 200 200 200
250 250 500 500 250 250 500 500 375 375 164.7 585.2 375 375 375 375 375 375 375 375
60 60 60 60 180 180 180 180 120 120 120 120 19.09 220.9 120 120 120 120 120 120
Observed % Y
Predicted % Y
7.54 61.71 14.10 70.66 30.52 68.40 51.80 85.60 7.27 78.41 47.94 74.65 16.34 73.31 67.81 67.50 67.64 68.14 67.73 67.97
4.07 58.49 13.24 66.81 34.18 69.07 54.83 88.88 5.78 80.17 49.25 73.61 23.02 66.89 67.79 67.79 67.79 67.79 67.79 67.97
12
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
Table 3 Experimental range and levels of the independent variables. Variables
Factor
Temperature ( C) Oxidant (H2O2, mM) Time (minutes)
x1 x2 x3
Range and level a
1
0
þ1
þa
31.82 164.77 19.09
100 250 60
200 375 120
300 500 180
368.17 585.22 220.90
data, the levels of the three main parameters investigated in this study are presented in Table 3. For statistical calculations, the variables Xi (the real value of an independent variable) were coded as xi (dimensionless value of independent variable) according to the following equation:
xi ¼
ðXi Xo Þ dX
(2)
where, Xo is the value of Xi at the center point, and dX represents the €zmen, 2012). step change (Kayan and Go 3. Results and discussion 3.1. Optimization of the degradation process The effects of three parameters, namely temperature, H2O2 concentration and experimental time of degradation were studied with the help of Design-Expert and the subsequent statistical analysis was performed by RSM. A statistical approach with a central composite design was used for determining the interaction between these factors. A total of 20 experiments were conducted for three factors at five levels (Table 3) and the COD removal percentage was chosen as response. The effect of pressure was not studied because at temperatures below 300 C water is fairly incompressible, which means that pressure has little effect on the physical properties of water, provided it is sufficient to maintain the liquid state (Ramos et al., 2002; Carr et al., 2011). 3.2. Multiple regression modeling The data in Table 2 were used to fit the polynomial model representing the degradation percentage (response) as a function of temperature, H2O2 concentration and experimental time. According to the RSM results, polynomial regression modeling was operated between the responses of the corresponding coded values of the three different process variables, and finally the optimum fit €zmen, model equation was obtained as follows (Kayan and Go 2012):
Y ¼ 67:79 þ 22:12x1 þ 7:24x2 þ 13:04x3 0:21x1 x2 4:88x1 x3 þ 2:87x2 x3 8:77x21 2:25x22 8:07x23
(3)
where Y is the response show as the predicted COD percentage. x1, x2 and x3 are the corresponding coded variables of temperature, H2O2 concentration, and experimental time, respectively. 3.3. Analysis of variance (ANOVA) In order to ensure the statistical significance of the quadratic model employed for explaining the experimental data at a 95% confidence level, the model was tested by analysis of variance (ANOVA) results. The analysis of variance of regression parameters of the response surface methodology quadratic model for landfill leachate degradation in subcritical water are listed in
Table 4. On the basis of the experimental values, statistical testing was carried out using Fisher's test for ANOVA (Moghaddam et al., 2011). From Table 4 it was observed that the regression was statistically significant at an F-value of 79.36 for landfill leachate degradation with a very low probability value (Pmodel
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Subcritical water treatment of landfill leachate: Application of response surface methodology P. Kirmizakis a, C. Tsamoutsoglou b, B. Kayan c, D. Kalderis a, * a
Department of Environmental and Natural Resources Engineering, School of Applied Sciences, Technological and Educational Institute of Crete, Chania, Crete 73100, Greece b DEDISA S.A., Chania, Crete 73100, Greece c Department of Chemistry, Arts and Sciences Faculty, Aksaray University, Aksaray, Turkey
a r t i c l e i n f o
a b s t r a c t
Article history: Received 4 October 2013 Received in revised form 9 April 2014 Accepted 21 April 2014 Available online
Context: Leachate is the liquid formed when waste breaks down in the landfill and water filters through that waste. This liquid is highly toxic and can pollute the land, ground water and water ways. It is mandatory for landfills to protect against leachate in most countries worldwide. Controlling the pollutant loading, means reducing its quantity by containing or treating the waste to comply with certain discharge characteristics which are compatible with the receptor medium. Objective: This paper describes the reduction of the organic load of a mature landfill leachate using a novel experimental set-up that employs hydrogen peroxide under subcritical conditions and aims to establish this method as an effective alternative to currently used options. Response surface methodology was applied to optimize the treatment process and determine which of the following there parameters e temperature, residence time and hydrogen peroxide concentration e played the most important role. Method: The method employed is based on the use of laboratory-scale, stainless steel reactors, filled with the leachate and appropriate quantities of hydrogen peroxide. Under subcritical conditions (temperature in the range of 100e374 C and enough pressure to maintain the liquid state of water), hydrogen peroxide produces hydroxyl radicals which are highly reactive and oxidize the organic molecules of the leachate. Results: The highest COD decrease of 85% was experimentally observed at 300 C, 500 mM H2O2 and 180 min residence time. It was determined that the combination of oxidant concentration and temperature is the rate-determining factor, whereas residence time has a lesser effect on the process. Conclusions: A simple, quick, effective and environmentally-friendly method for the treatment of the organic load of landfill leachate was developed and optimized at laboratory scale. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Landfill leachate Subcritical water Superheated water Response surface methodology Central composite design Hydrogen peroxide
1. Introduction Landfill leachate can be defined as the liquid produced from the decomposition of solid waste and infiltration of rainwater in a landfill. It contains heavy metals, salts, nitrogen compounds and various types of organic substances (Schiopu and Gavrilescu, 2010). Generation of leachate occurs when moisture enters the solid waste in a landfill, dissolves the contaminants into liquid phase and
* Corresponding author. Tel.: þ30 2821023017; fax: þ30 2821023003. E-mail addresses: [email protected], [email protected] (D. Kalderis). http://dx.doi.org/10.1016/j.jenvman.2014.04.037 0301-4797/© 2014 Elsevier Ltd. All rights reserved.
produces moisture content sufficient to initiate liquid flow. This leachate is a high-polluting strength wastewater that has a major impact and influence on landfill design and its operation. Leachate composition varies from one landfill to another, and over space and time in a particular landfill with fluctuations that depend on short and long-term periods due to variations in climate, hydrogeology and waste composition (Keenan et al., 1984). Generally, leachate possesses high concentrations of ammonia and organic contaminants (measured in terms of chemical oxygen demand COD and biochemical oxygen demand BOD), halogenated hydrocarbons and heavy metals. In addition, leachate usually contains high concentrations of inorganic salts (mainly sodium chloride, carbonate and sulphate) and is dependent on the
10
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
composition of landfilled waste. The main environmental problems at landfill sites are the infiltration of leachate and its subsequent contamination of the surrounding land and aquifers. Improvements in landfill engineering are aimed at reducing leachate production, collection and treatment prior to discharge (Farquhar, 1989). The risks to human health and the environmental impact of landfill leachate contamination of ground- and surface water have been well documented (Slack et al., 2005; Macklin et al., 2011; Kwasniewska et al., 2012; Toufexi et al., 2013). An excess of 200 organic compounds have been identified in municipal landfill leachate, with more than 35 of them having the potential to cause harm to the environment and human health (Slack et al., 2005). Leachate needs to be pre-treated on site to meet the standards for its discharge into the sewer or its direct disposal into surface water. Leachate management is a complex task due to the highly variable nature of waste landfilled, type and design of the landfill site, its age, and climatic and seasonal variations in different regions. Hence, it is difficult to recommend treatment options merely based on leachate age but it is often necessary to consider each case individually. Treatment systems in recent years are sophisticated, reliable and able to consistently treat leachate to keep in line with specific discharge standards. Generally, high organic and ammonia loads are the key factors in leachate treatment. The feasibility of treating leachate to COD levels lower than 1000 mg/L is uncertain, since at these values COD is primarily composed of humic and fulvic acids. The effect of these substances on aquatic life of receiving waters is dependent extensively on specific cases. While developing a treatment sequence for leachate to be discharged into a river, Robinson et al. (2002) reported that COD levels of about 500 mg/L consisting of humic and fulvic acids did not adversely affect aquatic life. Worldwide, the issue of leachate treatment has been thoroughly investigated, but a universal solution has not yet been established. Therefore, there is a need to develop reliable and sustainable options to manage leachate generation and treatment effectively. While designing a treatment system, the process train must include techniques or unit operations to treat leachate produced from the landfill over a longer period. The most widely used treatment processes and their combinations have been thoroughly reviewed (Kjeldsen et al., 2002; Renou et al., 2008; Abbas et al., 2009; Foo and Hameed, 2009; Schiopu and Gavrilescu, 2010; Li et al., 2010; Oller et al., 2011; Ahmed et al., 2012). Subcritical water is hot water (>100 C) under enough pressure to maintain the liquid state. It is an environmentally friendly and inexpensive solvent that exhibits a wide range of properties that render it very effective in solvating and decomposing moderately polar or non-polar organic substances from a wide range of environmental matrices. Several studies have shown that subcritical water can decompose naturally-occurring substances and materials, such as complex amino acids, proteins and carbohydrates (sucrose, fructose, sorbose) and brown coal to produce more valuable and useful products. Additionally, subcritical water has been proven to decompose hazardous organic substances and materials such as residual reactive dyes (Daskalaki et al., 2011), fluorochemicals (Hori et al., 2008), explosives (Hawthorne et al., 2000; Kalderis et al., 2008), pentachlorophenol (Benedictus, 2007), dioxins (Hashimoto et al., 2004) and polyvinyl chloride (Takeshita et al., 2004). The scope of this work is to reduce the organic load of a mature landfill leachate using a novel experimental set-up that employs hydrogen peroxide under subcritical conditions and therefore establish this method as an effective alternative to currently used options. Using subcritical water, with or without oxidants is an effective way of degrading various environmental pollutants. Such
subcritical water set-up has not been used for the treatment of landfill leachate before. Compared to the studies mentioned earlier, this time the process aims at a highly complex wastewater consisting of a large number of organic and inorganic constituents of diverse nature. Hydrogen peroxide is used as an environmentallyfriendly oxidant, as it leaves no residues after treatment. In contrast to the dynamic set-ups that require considerable quantities of water and more complicated apparatus, the experiments described here are under static conditions i.e. no flow is required and no additional use of water. Additionally, monitoring of the process is not essential, since the oven can be pre-set at the required temperature and residence time. Finally, no pumping system is needed to maintain the system pressure, since pressure is automatically controlled by the steam/water equilibrium inside the reactor cell. When applying oxidative methods to decrease the organic load of the leachate, temperature, oxidant concentration and experimental time are the most important factors. Therefore, optimization of the influencing parameters is vital towards designing an effective degradation system. As a result, advanced statistical design has been widely employed for process characterization, optimization and modeling (Alaton et al., 2009; Singh et al., 2010). Experimental design methodologies have become a significant tool that have enabled us to gain a better understanding of a process in terms of the interactions among the parameters that need to be optimized. Response surface methodology (RSM) e a collection of mathematical and statistical techniques e has been found to be a useful aid for studying the effect of several factors that influence the response of a system as well as optimizing the variables of a wastewater treatment process. Furthermore, it is essential to choose an appropriate experimental design method that will evaluate the effects of the major parameters involved in the treatment method and their probable interactions, through the minimum number of experiments. RSM provides a large amount of information and is a more economical approach because a small number of experiments are performed for monitoring the interaction of the independent variables on the response. In conventional optimization, the increase in the number of experiments necessary to carry out the research, leads to an increase in time and expenses as well as an increase in the utilization of reagents and materials for experiments. Based on the above, the objectives of this study are the following: effectively reduce the organic load e as shown in COD measurements e of a mature landfill leachate using hydrogen peroxide in subcritical conditions apply a 5-level central composite experimental design (CCD) combined with RSM to optimize the various parameters and obtain maximum response, validate the suggested model and determine the optimum operational conditions, based on COD measurements
2. Materials and methods 2.1. Materials Landfill leachate was obtained from the Chania (Crete, Greece) municipal waste landfilling site with the help of the staff of DEDISA A.E. The sample was followed by an analytical certificate showing the levels of organic and inorganic constituents and other quality parameters (Table 1). The addition of the appropriate quantities of hydrogen peroxide (30% solution, SigmaeAldrich) was carried out
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
experiments were performed at the temperatures and residence times indicated by the central composite design (Table 2).
Table 1 Qualitative parameters of the landfill leachate sample obtained from DEDISA S.A. and other leachates described in literature. Parameter
Value
pH Conductivity (mS/cm1) Dissolved oxygen (mg L1) Total suspended solids (TSS, mg L1) NeNO3 (mg L1) NeNO2 (mg L1) 1 NeNHþ 4 (mg L ) P-PO4 (mg L1) TP (mg L1) BOD (mg L1) COD (mg L1) TOC (mg L1) 1 HCO 3 (mg L ) 1 SO2 4 (mg L ) Cl (mg L1) Naþ (mg L1) 1
This work
Alkassasbeh et al., 2009
8.1 22.5 3.4 447.50 16.76 0.50 2786.5 20.3 24.8 1434.4 12,360 698.3 10,676.7 8.5 2079.1 1644
7.4 26.3 0.78 nr1 nr nr 956 nr nr 1228 7600 nr nr nr nr nr
11
2.2. Methods 2.2.1. Central composite design and landfill leachate degradation optimization The design, mathematical modeling and optimization of this study were performed using Design Expert 8.0.5 software. Central composite design was used to model the RSM in this design. It is the most widely used experimental design for fitting a second-order response surface (Montgomery, 2009). It was first reported by Box and Wilson in 1951 and is well suited for fitting a quadratic surface, which usually works well for the process optimization € (Ozer et al., 2009). The independent variables (factors) used in this experimental study were temperature, H2O2 concentration and experimental time and are coded as x1, x2, and x3 respectively (Table 3). The CCD was conducted with 20 experiments, including 8 star points, 6 axial points corresponding to the alpha value and 6 replicates at the center points (Ghafari et al., 2009; Adlan et al., 2011). In the present study, such design was employed for determining the optimum conditions for landfill leachate organics degradation in subcritical water. The experimental range and levels of independent variables for landfill leachate degradation are given in Table 3. The experimental results were analyzed using Design Expert 8.0.5 and the regression model was proposed. In the optimization process, the responses can be simply related to the chosen factors by linear or quadratic models. A quadratic model, which also includes the linear model, is given below as Eq. (1).
7.6 14.7 2.52 nr nr nr 470 nr nr 866 3733 nr nr nr nr nr
not reported.
just before the experiment in order to avoid H2O2 photodecomposition. The experimental set-up and procedure is described in detail in Kalderis et al., 2008 and Daskalaki et al., 2011. Briefly, one type of small laboratory reactor was used for the degradation studies. The 25 ml reactors were constructed from (6-inches long, 0.64 inches i.d.) 316 stainless steel pipe with male national pipe threads (npt) and female end caps sealed with Teflon tape (Swagelok Company, USA). Each reactor was loaded with 20 ml landfill leachate, the appropriate quantity of H2O2 was added (according to the central composite design experiments e Table 2) and capped. This procedure left ca. 5 ml of headspace in the cell. All static (non-flowing) reaction cells must contain a sufficient headspace so that the pressure inside the cell is controlled by the steam/liquid equilibrium. A full cell must never be used since the pressures could reach several thousand bar. Cells were placed in a (pre-heated at the required temperature) GC oven (HewlettePackard 5890, series II) for heating. All
% Y ¼ bo þ
X
bi xi þ
X
bii xi2 þ
X
bij xi xj þ ε
(1)
where, bo is the constant coefficient, bi, bii and bij are the regression coefficient and xi, xj indicate the independent variables. ε represents the random error. The quality of the polynomial model was expressed by the coefficient of determination, namely, R2 and R2adj. The statistical significance was verified with adequate precision ratio and by the F-test (Rauf et al., 2008). According to the obtained experimental
Table 2 Central composite design experiments and experimental results. Experiment number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Experimental design
Experimental plan
T ( C)
H2O2(mM)
t (min)
x1
x2
x3
1 þ1 1 þ1 1 þ1 1 þ1 1.682 þ1.682 0 0 0 0 0 0 0 0 0 0
1 1 þ1 þ1 1 1 þ1 þ1 0 0 1.682 þ1.682 0 0 0 0 0 0 0 0
1 1 1 1 þ1 þ1 þ1 þ1 0 0 0 0 1.682 þ1.682 0 0 0 0 0 0
100 300 100 300 100 300 100 300 31.82 368.18 200 200 200 200 200 200 200 200 200 200
250 250 500 500 250 250 500 500 375 375 164.7 585.2 375 375 375 375 375 375 375 375
60 60 60 60 180 180 180 180 120 120 120 120 19.09 220.9 120 120 120 120 120 120
Observed % Y
Predicted % Y
7.54 61.71 14.10 70.66 30.52 68.40 51.80 85.60 7.27 78.41 47.94 74.65 16.34 73.31 67.81 67.50 67.64 68.14 67.73 67.97
4.07 58.49 13.24 66.81 34.18 69.07 54.83 88.88 5.78 80.17 49.25 73.61 23.02 66.89 67.79 67.79 67.79 67.79 67.79 67.97
12
P. Kirmizakis et al. / Journal of Environmental Management 146 (2014) 9e15
Table 3 Experimental range and levels of the independent variables. Variables
Factor
Temperature ( C) Oxidant (H2O2, mM) Time (minutes)
x1 x2 x3
Range and level a
1
0
þ1
þa
31.82 164.77 19.09
100 250 60
200 375 120
300 500 180
368.17 585.22 220.90
data, the levels of the three main parameters investigated in this study are presented in Table 3. For statistical calculations, the variables Xi (the real value of an independent variable) were coded as xi (dimensionless value of independent variable) according to the following equation:
xi ¼
ðXi Xo Þ dX
(2)
where, Xo is the value of Xi at the center point, and dX represents the €zmen, 2012). step change (Kayan and Go 3. Results and discussion 3.1. Optimization of the degradation process The effects of three parameters, namely temperature, H2O2 concentration and experimental time of degradation were studied with the help of Design-Expert and the subsequent statistical analysis was performed by RSM. A statistical approach with a central composite design was used for determining the interaction between these factors. A total of 20 experiments were conducted for three factors at five levels (Table 3) and the COD removal percentage was chosen as response. The effect of pressure was not studied because at temperatures below 300 C water is fairly incompressible, which means that pressure has little effect on the physical properties of water, provided it is sufficient to maintain the liquid state (Ramos et al., 2002; Carr et al., 2011). 3.2. Multiple regression modeling The data in Table 2 were used to fit the polynomial model representing the degradation percentage (response) as a function of temperature, H2O2 concentration and experimental time. According to the RSM results, polynomial regression modeling was operated between the responses of the corresponding coded values of the three different process variables, and finally the optimum fit €zmen, model equation was obtained as follows (Kayan and Go 2012):
Y ¼ 67:79 þ 22:12x1 þ 7:24x2 þ 13:04x3 0:21x1 x2 4:88x1 x3 þ 2:87x2 x3 8:77x21 2:25x22 8:07x23
(3)
where Y is the response show as the predicted COD percentage. x1, x2 and x3 are the corresponding coded variables of temperature, H2O2 concentration, and experimental time, respectively. 3.3. Analysis of variance (ANOVA) In order to ensure the statistical significance of the quadratic model employed for explaining the experimental data at a 95% confidence level, the model was tested by analysis of variance (ANOVA) results. The analysis of variance of regression parameters of the response surface methodology quadratic model for landfill leachate degradation in subcritical water are listed in
Table 4. On the basis of the experimental values, statistical testing was carried out using Fisher's test for ANOVA (Moghaddam et al., 2011). From Table 4 it was observed that the regression was statistically significant at an F-value of 79.36 for landfill leachate degradation with a very low probability value (Pmodel
