- Email: [email protected]
Optimization of phosphate removal from wastewater by electrocoagulation with aluminum plate electrodes
Separation and Purification Technology 52 (2006) 394–401
Optimization of phosphate removal from wastewater by electrocoagulation with aluminum plate electrodes S¸ahset ˙Irdemez, Yalc¸ın S¸evki Yildiz ∗ , Vahdettin Tosuno˘glu Department of Environmental Engineering, Atat¨urk University, 25240 Erzurum, Turkey Received 9 March 2006; received in revised form 18 May 2006; accepted 20 May 2006
Abstract The Taguchi method was used to determine the optimum conditions for the phosphate removal from wastewater by electrocoagulation with aluminum plate electrodes. The experimental parameters investigated were initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density. The ranges of experimental parameters were between 50 and 500 mg/L (as PO4 –P), 4–7 for initial pH, 0–10 mM, NaCl, NaNO3 , Na2 SO4 and CaCl2 and 0.25–1.00 mA/cm2 mm for initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density, respectively. Reaction period was kept constant in 25 min for statistical analysis. The optimum conditions for these parameters were found to be 50 mg/L, 4, 5 mM, NaCl and 1.00 mA/cm2 , respectively. Under these conditions, the predicted and experimental removal efficiency of phosphate from wastewater by electrocoagulation with aluminum plate electrodes were 99.9 and 100.0%, respectively. A statistical analysis of variance (ANOVA) was performed to see whether the process parameters were statistically significant or not. According to the F-test results, it can be concluded that the degrees of the influences of parameters on the removal efficiency is initial phosphate concentration, current density and initial pH of the solution. © 2006 Elsevier B.V. All rights reserved. Keywords: Taguchi method; Optimization; Phosphate removal; Electrocoagulation; Aluminum electrode
1. Introduction As well known, eutrophication is one of the main problems nowadays encountered in the monitoring of the environmental water sources the industrialized countries. This phenomenon is caused by the excess phosphorus concentration in the effluents from municipal or industrial plants discharged in the environment [1] the usual forms of phosphorus found in solutions include orthophosphate, polyphosphate and organic phosphate [2]. The principal phosphorus compounds in wastewater are generally orthophosphate forms together with smaller amounts of organic phosphate [3]. In the countryside, where agriculture and animal husbandry are the main industries, wastes from these activities will contribute to the accumulation of P in soil and water bodies. These phosphorus compounds, dissolved in surface or ground waters, are responsible for the eutrophication in closed water systems, especially in lakes and enclosed bays where the water is almost stagnant [4]. Phosphorus removal tech-
∗
Corresponding author. Tel.: +90 442 2314799; fax: +90 442 2360957. E-mail address: [email protected] (Y.S¸. Yildiz).
1383-5866/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2006.05.020
niques are chemical treatments like adsorption, chemical precipitation, ion exchange, electrodialysis, hybrid systems containing fly-ash adsorption and membrane filtration and electrocoagulation. Adsorption and chemical precipitation among the above methods have been widely used for phosphate removal [5–13]. The removal of phosphate from aqueous streams consists of the conversion of soluble phosphate to an insoluble solid phase. This solid phase can be separated from water by means of sedimentation or filtration. In wastewater applications, the most common and successful methods to precipitate phosphate involve the dissolved cations Al3+ , Ca2+ , Fe3+ and to a lesser extent of Fe2+ . It was found that when iron and aluminum are present in the water, FePO4 and AlPO4 forms in the low pH range (<6.5) and at higher pH range (>6.5) iron and aluminum increasingly convert to oxides and hydroxides. A higher pH is more ideal for precipitation of phosphate with calcium as apatites and hydroxyapatites [3]. In recent years, electrocoagulation has been successfully tested to treat wastewater. Electrocoagulation is a process consisting of creating metallic hydroxide flocks within the wastewater by electrodissolution of soluble anodes, usually made of iron or aluminum [14]. The difference between electro-
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
395
coagulation and chemical coagulation is mainly in the way of aluminum ions are delivered. In electrocoagulation, coagulation and precipitation are not conducted by delivering chemicals – called coagulants – to the system, but via electrodes in the reactor [14]. Electrocoagulation is based on the fact that the stability of colloids, suspensions and emulsions is influenced by electric charges. Therefore, if additional electrical charges are supplied to the charged particles via appropriate electrodes, the surface charge of particles is neutralized and several particles combine into larger and separable agglomerates [15]. Electrode assembly is the heart of the treatment facility. Therefore, the appropriate selection of its materials is very important. The most common electrode materials for electrocoagulation are aluminum and iron. They are cheap, readily available, and proven effective [16]. When aluminum is used as electrode material, the reactions are as follows: • At the cathode: 3H2 O + 3e− → 23 H2 (g) + 3OH−
Fig. 1. Schematic diagram of the experimental setup.
(1)
• At the anode: Al → Al3+ + 3e−
(2)
• In the solution: Al3+ (aq) + 3H2 O → Al(OH)3 + 3H+ (aq)
(3)
Taguchi’s orthogonal array (OA) analysis is used to obtain the best parameters for the optimum process design with the least number of experiments. In recent years, the Taguchi method has been used to determine optimum parameters because of its many advantages [17]. The main advantages of this method over other statistical experimental design methods are that the parameters affecting an experiment can be investigated as controlling and not controlling and that the method can be applied to an experimental design involving a large number of design factors [18,19]. Aim of this study is not to investigate the treatability of the phosphate containing wastewater by electrocoagulation method. Our aim is to determine the optimum operating conditions such as initial phosphate concentration, supporting electrolyte type and concentration, current density and initial pH of the wastewater for the removal of phosphate from waters by electrocoagulation method with plate aluminum electrodes based on removal efficiency.
filter with the pore diameter of 045 m (Schleicher and Schuell) before the analysis. The analysis of phosphate was carried out using the yellow vanadomolybdophosphoric acid method by a double beam spectrophotometer (Shimadzu UV-160 A) according to the Standard Methods for Examination of Water and Wastewater [20]. The initial pH was adjusted to a desired value using NaOH (Merck, 5N) or HNO3 (Carlo Erbaa, 65%). 2.2. Experimental setup and procedure
2. Experimental
The experimental setup is schematically shown in Fig. 1. The electrocoagulation unit consists of six pair of electrodes made of plate aluminum with total area of approximately 1500 cm2 and the gap between the electrodes is 5 mm. Electrodes were connected to a digital dc power supply (Shenzen-Mastech HY 3005-3) in monopolar mode. Two digital multimeters (Brymen Bm 201) as ampermeter and voltmeter were used to measure the current passing through the circuit and the applied potential, respectively. The electrocoagulation unit has been stirred at 150 rpm by a magnetic stirrer (Heidolp MR 3004 S). The experimental setup is shown in Fig. 1. The thermostated electrocoagulator is made of plexiglass with the volume of 850 mL. During the experiments, temperature, conductivity and pH of the wastewaters were measured by a multi-parameter (WTW Multiline P-4 F-Set-3). Reactor was operated in batch and galvanostatic mode.
2.1. Materials
2.3. Statistical analysis
All chemicals used were analytical grade and used without any further treatment. Distilled water was used in all experiments. Phosphate solutions were prepared from KH2 PO4 (Riedel de Ha¨en, 98%). NaCl (Merck, 99.5%), NaNO3 (Merck, 99%), Na2 SO4 (Sigma–Aldrich, 99%) and CaCl2 (Merck, >90) were used as supporting electrolyte. Treated wastewater was collected over a desired period of time from the reactor and collected samples were filtered by the cellulose acetate membrane
The variables chosen for this investigation are supporting electrolyte type and concentration, current density, initial phosphate concentration, and initial pH of the wastewater. The variables investigated and their levels were summarized in Table 1. Reaction period was kept constant in 25 min for statistical analysis. The use of the parameter design in the Taguchi method to optimize a process with multiple performance characteristics
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
396
Table 1 Variables and their values corresponding to their levels investigated in the experiments Variables
Levels
(A) Initial phosphate concentration (PO4 –P mg/L) (B) Initial pH of the wastewater (C) Supporting electrolyte concentration (mM) (D) Supporting electrolyte type (E) Current density (mA/cm2 )
1
2
3
4
50 4 0.0 NaCl 0.25
100 5 2.5 NaNO3 0.50
200 6 5.0 Na2 SO4 0.75
500 7 10.0 CaCl2 1.00
includes the following steps: (1) Identification of the performance characteristics and selection of the process parameters to be evaluated. (2) Determination of the number of parameter levels for the process and possible interaction between the process parameters. (3) Selection of the appropriate orthogonal array and assignment of process parameters to the orthogonal array. (4) Conduction of the experiments based on the arrangement of the orthogonal array. (5) Calculation of the performance characteristics. (6) Analysis of the experimental results by using the performance characteristics and ANOVA. (7) Selection of the optimal levels of process parameters. (8) Verification of the optimal process parameters through the confirmation experiment [21,18]. The orthogonal array (OA) experimental design was chosen as the most suitable method to determine an experimental plan, L16 (45 ) (Table 2), for five parameters each with four values [21]. The experimental variables, their levels and results of conducted experiments are given in Table 2. In order to observe the effects of noise sources on the electrocoagulation process, each experiment was repeated twice under
the same conditions at different times. The performance characteristics were chosen as the optimization criteria. There are three categories of performance characteristics: the larger the better – this category was used to evaluate the system performance based on phosphate removal efficiency – the smaller the better and the nominal the better. The two performance characteristics were evaluated by using Eqs. (4) and (5) [21]: • Larger is better
SNL = −10 log10
n
1 1 n Yi2
(4)
i=1
• Smaller is better SNS = −10 log10
n
1 2 Yi n
(5)
i=1
where SNL and SNS are the performance characteristics, n the number of repetitions done for an experimental combination and Yi is the performance value of the ith experiment. In the Taguchi method, the experiment corresponding to optimum working conditions might not have been done during the whole period of the experimental stage. In such cases, the performance value corresponding to the optimum working conditions can be predicted by utilizing the balanced characteristic of OA. For
Table 2 Experimental variables, their levels and results of conducted experiments corresponding to L16 experimental plan Experiment no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Variables and their levels
Removal efficiency of phosphate (%)
(A)
(B)
(C)
(D)
(E)
First series
Second series
Average
1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1 2 3 4 2 1 4 3 3 4 1 2 4 3 2 1
1 2 3 4 3 4 1 2 4 3 2 1 2 1 4 3
1 2 3 4 4 3 2 1 2 1 4 3 3 4 1 2
99.3 100.0 96.9 100.0 97.2 97.0 95.2 78.0 100.0 48.0 86.6 71.0 64.7 73.0 37.0 26.2
93.3 100.0 98.6 100.0 95.6 97.8 95.7 72.0 99.7 63.0 85.5 67.0 62.8 72.0 26.3 22.3
96.3 100.0 97.7 100.0 96.4 97.4 95.5 75.0 99.9 55.7 86.1 69.0 63.7 72.5 31.7 24.2
Energy consumption (kW h/m3 ), average 2.86 1.91 22.07 2.27 7.03 20.96 0.89 0.80 0.83 5.56 11.42 2.60 3.20 4.12 0.41 19.89
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
397
Table 3 Results of the analysis of variance for the removal efficiencies of phosphate Variables
Sum of squares
Degrees of freedom
Mean of squares
F
(A) Initial phosphate concentration (PO4 –P mg/L) (B) Initial pH of the wastewater (–) (C) Supporting electrolyte concentration (mM) (D) Supporting electrolyte type (–) (E) Current density (mA/cm2 ) Error
11895.6 2008.3 674.5 1156.7 2489.8 228.2
3 3 3 3 3 16
3965.2 669.4 224.8 385.6 829.9 14.3
278.08 46.95 15.77 27.09 58.20
this, the following additive model may be used: Yi = m + X i + e i
(6)
where m is the overall mean of performance value, Xi the fixed effect of the parameter level combination used in the ith experiment and ei is the random error in the ith experiment. If experimental results are stated in a percentage (%), before evaluating Eq. (6), the Ω transformation of percentage values should be applied first using Eq. (7) by which values of interest are also determined later by carrying out a reverse transformation by using the same equation [22]: 1 Ω(db) = −10 log −1 (7) P where Ω(db) is the decibel value of percentage value subject to omega transformation and P is the percentage of the product obtained experimentally. Since Eq. (6) is a point estimation which is calculated by using experimental data in order to determine whether the additive model is adequate or not, the confidence limits for the prediction error must be evaluated [21]. The prediction error is the difference between the observed Yi and the predicted Yi . The confidence limits for the prediction error are 1 1 2 σ + σ2 Se = ±2 (8) n0 e nr e sum of squares due to error (9) degrees of freedom for error 1 1 1 1 1 1 1 1 + + + ··· = + − − − n0 n nAi n nBi n nCi n (10)
σe2 =
where Se is the two-standard deviation confidence limit, n the number of rows in the matrix experiment, nr the number of repetitions in the confirmation experiment and nAi , nBi , nCi , . . . are the replication numbers for the parameter levels Ai, Bi, Ci, . . . If the prediction error is outside these limits, the possibility that the additive model is not adequate should be suspected. Otherwise, the additive model can be considered to be adequate. A verification experiment is a powerful tool for detecting the presence of interactions among the control parameters. If the predicted response under the optimum conditions does not match the observed response, it implies that the interactions are important. If the predicted response matches the observed response, it then implies that the interactions are probably not important and that the additive model is a good approximation [29]. The order of the experiments was obtained by inserting
parameters into the columns of OA and L16 (45 ) which were chosen as the experimental plan given in Table 2. The order of experiments was made random in order to avoid noise sources which had not been considered initially and which could take place during an experiment and affect results in a negative way. 3. Results and discussions 3.1. Statistical analysis The collected data were analyzed by an IBM compatible PC. In order to see effective parameters and their confidence levels on the electrocoagulation process, the analysis of variance was performed. A statistical analysis of variance (ANOVA) was performed to see whether the process parameters were statistically significant or not. The F-test is a tool to see which process parameters have a significant effect on the removal efficiency. The F-value for each process parameter is simply a ratio of mean of the squared deviations to the mean of squared error. Usually, the larger the F-value, the greater the effect on the performance criteria value due to the change of the process parameter. With the performance characteristics and ANOVA analyses, the optimal combination of process parameters can be predicted [17]. The results of variance analysis are given in Table 3. F values in the table show that the effective parameters on the removal of phosphate by electrocoagulation method are initial phosphate concentration, current density, initial pH of the wastewater and supporting electrolyte type. To obtain optimal phosphate removal performance, the larger the better performance characteristic (Eq. (4)) has been taken for the removal of phosphate. The optimal level of a process parameter is the level with the highest SN value calculated by Eq. (4). Fig. 2 shows the variation of the performance characteristics with the variables. To determine the experimental conditions for the first data point, the CD
Fig. 2. The effect of each parameter on the optimization criteria for phosphate removal.
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
398
Table 4 Optimum working conditions and alternative working conditions for different experimental setups observed and predicted removal efficiencies of phosphate Case no.
1* 2 3 4 5 6 7 8 9 10 11 *
A (Cf )
B (pHi )
OBS.
PRED
CONF
Level
Value
Level
Value
C (sec) Level
Value
D (set) Level
Value
E (CD) Level
Value
–
–
–
1 1 1 1 1 2 1 2 2 2 3
50 50 50 50 50 100 50 100 100 100 200
1 1 1 1 1 1 1 4 3 1 1
4 4 4 4 4 4 4 7 6 4 4
3 4 3 3 4 2 4 3 2 2 2
5 10 5 5 10 2.5 10 5 2.5 2.5 2.5
1 4 4 2 2 3 1 1 4 1 1
NaCl CaCl2 CaCl2 NaNO3 NaNO3 Na2 SO4 NaCl NaCl CaCl2 NaCl NaCl
4 4 4 4 4 4 1 2 1 1 1
1 1 1 1 1 1 0.25 0.50 0.25 0.25 0.25
100.0 100.0 100.0 98.0 98.2 96.9 93.9 88.0 99.5 87.0 63.0
99.9 100.0 100.0 100.0 100.0 96.4 99.4 84.5 76.0 87.3 76.8
88.9–100.0 89.0–100.0 89.0–100.0 89.0–100.0 89.0–100.0 85.4–100.0 88.34–100.0 73.5–100.0 65.0–87.5 76.3–98.4 65.8–87.8
Optimum working conditions
(current density) for that point is level 1 which is 0.25 mA/cm2 for this parameter. The experiments for which CD level is 1 are experiments 1, 8, 10 and 15. The performance characteristics value of the first data point is thus the average of those obtained from experiments 1, 8, 10 and 15. Thus, experimental conditions for the fourth data point are the conditions of the experiments for which column CD is 2 (experiments 4, 5, 11 and 14) [23,24]. The numerical value of the maximum point in each graph marked the best value of that particular parameter and was found as A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and E4(1 mA/cm2 ). These parameter values provide the optimum conditions. If experimental plan given in Table 2 is studied carefully together with parameter values given A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and CD4(1 mA/cm2 ) it can be seen that experiments corresponding to optimum conditions A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and CD4(1 mA/cm2 ) have not been carried out during the experimental work. Thus, it should be noted that the removal efficiencies for phosphate (%) in Table 4 are predicted results obtained by using Eq. (6) and observed results for the same conditions. Also, the results in Table 4 are confidence limits of predictions. In order to test the predicted results, confirmation experiments were carried out once at the same working conditions. The fact that the removal efficiencies from the confirmation experiments are within the calculated confidence intervals calculated from Eqs. (8)–(10) (Table 4) shows that the experimental results are within ±5% in error. This case states that there is a good agreement between the predicted values and experimental values and the interactive effects between the parameters are indeed negligible. It may be concluded that the additive model is adequate for describing the dependence of the dissolution process on the various parameters [21].
the preceding equation can be integrated to give the following: C0 = K1 At (12) V ln Ct It can be concluded that Eq. (12) can be applied to removal of phosphate from waters by electrocoagulation. Fig. 3 illustrates the variation of removal efficiency (η) with time. Figs. 3 and 4 were plotted using the data obtained from experiment 3 and similar tendencies were observed in other test runs (not shown). Sharp increases of removal efficiencies are clearly observed initially. After the treatment period of 15 min the removal efficiencies approach plateaus at 90% (the point with arrow in Fig. 3). The slight improvement for phosphate removal following a great increase in the energy consumption (see Fig. 4) and is not worth the energy consumed because of the rapid increases in applied potential as shown in Fig. 4. Removal efficiencies and energy consumptions increases with increasing time, but increase in energy consumption is higher than that of removal efficiency beyond the plateau in Fig. 3. For example, system consumes electrical energy of 14.37 kW h/m3 and provides removal efficiency of 94.4% for 16 min while energy consumptions raise from 14.37 kW h/m3
3.2. Kinetic analysis of the electrocoagulation process According to Hosny, the rate of oil removal from oil–water emulsions can be expressed by Eq. (11) [25]: dC −V (11) = K1 AC dt
Fig. 3. Variation of phosphate removal efficiency with time (C0 = 50 mg/L, pHi = 6.0, CSE = 5.0 mM, Na2 SO4 , CD = 0.75 mA/cm2 ).
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
399
Table 5 Variation of removal rate constants with experimental variables corresponding to L16 experimental plan Experiment no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Variables and their levels (A)
(B)
(C)
(D)
(E)
1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1 2 3 4 2 1 4 3 3 4 1 2 4 3 2 1
1 2 3 4 3 4 1 2 4 3 2 1 2 1 4 3
1 2 3 4 4 3 2 1 2 1 4 3 3 4 1 2
η (%), average
t (min), average
K1 (cm/min), average
K1 /I (×10,000 cm/A-min), average
94.4 85.0 85.5 91.8 96.4 93.4 92.7 77.1 98.1 76.5 93.4 87.4 77.6 89.2 67.8 54.6
22 12 19 4 25 18 23 40 20 50 32 40 50 55 80 60
886.3 976.9 678.6 3935.8 733.8 883.7 663.0 227.8 916.0 171.7 480.5 277.4 181.3 236.9 74.8 83.0
2363.4 1302.6 603.2 2623.9 489.2 785.5 884.0 607.5 1221.4 457.9 320.4 246.6 161.2 157.9 199.5 110.7
All calculations conducted for K1 values were performed based on the time for reach the beginning of the plateau and presented in Table 5. When examined Table 5, it can be seen that cell removal constant has been strongly affected by the current passing through the circuit. Hence, it is more favorable that Eq. (12) should be modified as noted Eq. (13): V ln
Fig. 4. Variation of specific energy consumption with time (C0 = 50 mg/L, pHi = 6.0, CSE = 5.0 mM, Na2 SO4 , CD = 0.75 mA/cm2 ).
to about 22 kW h/m3 removal efficiencies increase from 94.4% to only 98.6% between the 16 and 25 min. Thus the system should be operated at the end of the sharp increase step.
C0 Ct
= K1 AIt
(13)
Maximum values of cell removal constants are achievable, when CaCl2 (Merck, 90%) is used as supporting electrolyte (see Table 5, exp. nos. 4, 9 and 15). Cell removal constants calculated according to optimum and alternative working conditions have been presented in Table 6. It can be said that the tendency based on cell removal constants in Table 6 is the same as Table 5 (see Table 6, exp. nos. 2, 3 and 9). Cell removal constants calculated for CaCl2 have the maximum values within the groups arranged according to initial phosphate concentrations. For example, in the initial phosphate
Table 6 Cell removal constants calculated according to optimum and alternative working conditions Experiment no.
1* 2 3 4 5 6 7 8 9 10 11 *
Variables and their levels (A)
(B)
(C)
(D)
(E)
1 1 1 1 1 2 1 2 2 2 3
1 1 1 1 1 1 1 4 3 1 1
3 4 3 3 4 2 4 3 2 2 2
1 4 4 2 2 3 1 1 4 1 1
4 4 4 4 4 4 1 2 1 1 1
Optimum working conditions
η (%)
t (min)
96.6 97.9 97.3 91.2 91.1 96.0 89.6 85.0 96.2 89.8 83.9
10 4 4 10 6 8 16 20 16 28 50
K1 /I (×10,000 cm/A-min)
4630.5 12942.0 12814.5 3567.0 6526.5 6078.0 7710.0 3006.0 26106.0 24312.0 2220.0
400
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
Fig. 6. X-ray diffractogram of precipitate obtained from the experiments (100 mg/L PO4 –P).
Fig. 5. (a) Variation of wastewater pH with time (experiments 1, 5 and 6 are confirmation test runs presented in Table 6); (b) variation of wastewater pH with time (experiments 2, 3 and 9 are confirmation test runs presented in Table 6).
concentration of 50 mg/L; cell removal constants for exp. nos. 2 and 3 equal to 12,942 and 12814.5 cm/A-min, respectively. Similarly 100 mg/L of initial phosphate concentration; cell removal constant is 26,106 cm/A-min for exp. no. 9. Hydroxyapatites, which are slightly soluble, form [26] according to reaction (14) [27] within the pH range of 4.0–8.5. When CaCl2 is used as supporting electrolyte: 5Ca2+ + 7OH− + 3H2 PO4 − → CaOH(PO4 )3 + 6H2 O
(14)
Variation of the wastewater pH with time has been presented in Fig. 5a. In all test runs which have been used as supporting electrolyte different from CaCl2 , waste water pH gradually increases with time. For example, within 20 min, pH values of the wastewater have increased approximately unit of 4.35, 4.60 and 4.00 for the experiment nos. 1, 5 and 6, respectively. Whereas wastewater pH is slightly increases with time when CaCl2 is used In the first 20 min, the pH values of the wastewater have increased unit of about 1, 2 and 1.2 for the experiment nos. 2, 3 and 9, respectively (see Fig. 5b). This behavior of the system can be attributed to the reaction (14) as seen in reaction (14), in order to form 1 mol of hydroxyapatite, 7 mol of OH− ions should be consumed, hence pH of the wastewater is not be able to increase. In order to determine the species in the precipitate, X-ray diffractograms have been used. The results of the X-ray
diffractogram of the precipitate obtained from the experiments conducted with containing100 mg/L PO4 –P have been given in Fig. 6. As seen in Fig. 6, Al(OH)3 and AlPO4 are available in the medium. The precipitate produced by the process of phosphate removal using electrocoagulation method is usually a mixture of Al(OH)3 and AlPO4 although the AlPO4 precipitation is favored over Al(OH)3 [27], when other ions are absence in the wastewater namely not using any supporting electrolyte. Preliminary investigations carried out by the authors have shown that aluminum electrodes provided the better results comparing to iron electrodes. Aluminum electrodes have higher removal efficiency and lower energy consumption than those of iron electrodes [28,29] and The effluent with aluminum electrodes was found very clear and stable, whereas the effluent with iron electrodes appeared greenish first, and then turned yellow and turbid. The green and yellow colors must have resulted from Fe(II) and Fe(III). Fe(II) is the common ion generated in situ of electrolysis of iron [16]. 4. Conclusions In this investigation, the Taguchi method was used to determine the optimum conditions for the phosphate removal. Effect of initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density on the electrocoagulation of phosphate ion has been investigated and effects of these parameters on the system performance have been evaluated based on removal efficiency. It can be said that the Taguchi method is able to use for the optimization of phosphate ions from wastewaters by electrocoagulation method: 1. The effective parameters on the removal of phosphate from wastewaters by electrocoagulation are initial phosphate concentration, current density and initial pH of the solution. 2. The optimum conditions for the parameters initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density were equal to 50 mg/L, 4, 5 mM NaCl and 1 mA/cm2 ,
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
respectively. Under these conditions, a removal efficiency of 100% can be achieved (Table 4). 3. It can be said that maximum cell removal constant values are achievable in the experiment groups arranged according to initial phosphate concentrations when CaCl2 is used as supporting electrolyte. Besides electrocoagulation system can reach the plateau in the shortest treatment periods (see Table 5), so detailed cost analysis of the system should be conducted and using potential of the CaCl2 as supporting electrolyte should be evaluated. 4. It can be concluded that the phosphate can be removed completely from water by using electrocoagulation. In electrocoagulation, the main reaction is phosphate removal has been accomplished accompanying with the precipitation of Al(OH)3 and AlPO4 although the AlPO4 precipitation is favored over Al(OH)3 and detailed economic analysis of the whole process is necessary for a more precise evaluation of the process (see Table 2 for energy consumptions). 5. The predicted and observed removal efficiency values are close to each other. Thus, it may be concluded that the additive model is adequate for describing the dependence of the removal efficiencies on the various parameters. Acknowledgements Authors are grateful to the research council of Atat¨urk University, for providing financial support with grant no. 2002/143. The authors also appreciate Dr. Ferhat B¨ulb¨ul for providing the valuable comments for evaluation of X-ray diffractograms. References [1] C. Sommariva, A. Converti, M. Del Borghi, Increase in phosphate removal from wastewater by alternating aerobic and anaerobic conditions, Desalination 108 (1996) 255–260. [2] G. Tchobanoglous, F.L. Burton, Wastewater Engineering, McGraw-Hill, 1991. [3] D.G. Grubb, M.S. Guimaraes, R. Valencia, Phosphate immobilization using an acidic type F fly ash, J. Hazard. Mater. 76 (2000) 217– 236. [4] S.H. Lee, W.H. Moon, Adsorption of phosphorus in saturated slag media columns, Sep. Purif. Technol. 12 (1997) 109–118. [5] L. Johansson, J.P. Gustafsson, Phosphate removal using blast furnace slags and opoka-mechanisms, Water Res. 34 (2000) 259. [6] N.M. Agyei, C.A. Strydom, J.H. Potgieter, An investigation of phosphate ion adsorption from aqueous solution by fly ash and slag, Cem. Concr. Res. 30 (2000) 823. [7] M. Rao, A.V. Parwate, A.G. Bhole, Removal of Cr6+ and Ni2+ from aqueous solution using bagasse and fly ash, Waste Manage. 22 (2002) 821–830.
401
[8] C.J. Lin, J.E. Chang, Effect of fly ash characteristics on the removal of Cu(II) from aqueous solution, Chemosphere 44 (5) (2001) 1185–1192. [9] B. Bayat, Combined removal of zinc(II) and cadmium(II) from aqueous solutions by adsorption onto high-calcium Turkish fly ash, Water Air Soil Pollut. 136 (2002) 69–92. [10] P.C. Kao, J.H. Tzeng, T.L. Huang, Removal of chlorophenols from aqueous solution by fly ash, J. Hazard. Mater. 76 (2000) 237–249. [11] S¸. ˙Irdemez, Y.S¸. Yıldız, N. Demircio˘glu, Phosphate removal from wastewaters by using fly-ash adsorption, in: Proceedings of the International Soil Congress, Erzurum, Turkey, 2004, pp. 29–34. [12] E. Yıldız, Phosphate removal from water by fly ash using crossflow microfiltration, Sep. Purif. Technol. 35 (2004) 241–252. [13] S¸. ˙Irdemez, Phosphate removal from wastewaters by electrocoagulation, Ph.D. Thesis, School of Natural and Applied Science, Atat¨urk University, Erzurum, 2005. [14] A.S. Koparal, The removal of salinity from produced formation water by conventional and electrochemical methods, Fresen. Environ. Bull. 11 (12a) (2002) 1071–1077. ¨ gu¨ tveren, S. Koparal, Electrocoagulation for oil-water em¨ulsion ¨ [15] U.B. O˘ treatment, J. Environ. Sci. Health A 32 (9/10) (1997) 2507–2520. [16] X. Chen, G.C. Chen, P.L. Yue, Separation of pollutants from restaurant wastewater by electrocoagulation, Sep. Purif. Technol. 19 (2000) 65–76. [17] C.Y. Nian, W.H. Yang, Y.S.J. Tarng, Optimization of turning operations with multiple performance characteristics, Mater. Process. Technol. 95 (1999) 90. ¨ K¨uc¸u¨ k, M.M. Kocakerim, M. C [18] O. ¸ opur, A. Yartas¸ı, Optimization of dissolution of ulexite in (NH4 )2 SO4 solutions, Can. Metall. Quart. 44 (2005) 1–6. ¨ K¨uc¸u¨ k, Application of Taguchi method in the optimization of dissolution [19] O. of ulexite in NH4 Cl solutions, Korean J. Chem. Eng. 23 (1) (2006) 21–27. [20] Standard Methods for Water and Wastewater Examination, 16th ed., AWWA, APHA, WPCF, New York, 1985. [21] M.S. Phadke, Quality Engineering Using Robust Design, Prentice Hall, New Jersey, 1989, pp. 61–292. [22] G. Taguchi, System of Experimental Design, Quality Resources, New York, 1987, p. 108. [23] M. C ¸ opur, T. Pekdemir, C. C ¸ elik, S. C ¸ olak, Determination of the optimum conditions for the dissolution of stibnite in HCl solutions, Ind. Eng. Chem. Res. 36 (3) (1997) 682–687. [24] M. C ¸ opur, An optimization study of dissolution of Zn and Cu in ZnS concentrate with HNO3 solutions, Chem. Biochem. Eng. 15 (4) (2002) 191–197. [25] A.Y. Hosny, Separating oil from oil-water emulsions by electroflotation technique, Sep. Technol. 6 (1996) 9–17. [26] W.J. Eilbeck, G. Mattock, Chemical Processes in Wastewater Treatment, John Wiley & Sons, New York, 1987. [27] W.W. Eckenfelder Jr., Industrial Water Pollution Control, 2nd ed., McGrawHill, 1989. [28] S¸. ˙Irdemez, N. Demircio˘glu, Y.S¸. Yıldız, The effects of pH on phosphate removal from wastewater by electrocoagulation with iron plate electrodes. J. Hazard. Mater., 2006, in press. [29] S¸. ˙Irdemez, N. Demircio˘glu, Y.S¸. Yıldız, Z. Bing¨ul, The effects of current density and phosphate concentration on phosphate removal from wastewater by electrocoagulation using aluminum and iron plate electrodes, Sep. Purif. Technol. 52 (2006) 218–223.
Optimization of phosphate removal from wastewater by electrocoagulation with aluminum plate electrodes S¸ahset ˙Irdemez, Yalc¸ın S¸evki Yildiz ∗ , Vahdettin Tosuno˘glu Department of Environmental Engineering, Atat¨urk University, 25240 Erzurum, Turkey Received 9 March 2006; received in revised form 18 May 2006; accepted 20 May 2006
Abstract The Taguchi method was used to determine the optimum conditions for the phosphate removal from wastewater by electrocoagulation with aluminum plate electrodes. The experimental parameters investigated were initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density. The ranges of experimental parameters were between 50 and 500 mg/L (as PO4 –P), 4–7 for initial pH, 0–10 mM, NaCl, NaNO3 , Na2 SO4 and CaCl2 and 0.25–1.00 mA/cm2 mm for initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density, respectively. Reaction period was kept constant in 25 min for statistical analysis. The optimum conditions for these parameters were found to be 50 mg/L, 4, 5 mM, NaCl and 1.00 mA/cm2 , respectively. Under these conditions, the predicted and experimental removal efficiency of phosphate from wastewater by electrocoagulation with aluminum plate electrodes were 99.9 and 100.0%, respectively. A statistical analysis of variance (ANOVA) was performed to see whether the process parameters were statistically significant or not. According to the F-test results, it can be concluded that the degrees of the influences of parameters on the removal efficiency is initial phosphate concentration, current density and initial pH of the solution. © 2006 Elsevier B.V. All rights reserved. Keywords: Taguchi method; Optimization; Phosphate removal; Electrocoagulation; Aluminum electrode
1. Introduction As well known, eutrophication is one of the main problems nowadays encountered in the monitoring of the environmental water sources the industrialized countries. This phenomenon is caused by the excess phosphorus concentration in the effluents from municipal or industrial plants discharged in the environment [1] the usual forms of phosphorus found in solutions include orthophosphate, polyphosphate and organic phosphate [2]. The principal phosphorus compounds in wastewater are generally orthophosphate forms together with smaller amounts of organic phosphate [3]. In the countryside, where agriculture and animal husbandry are the main industries, wastes from these activities will contribute to the accumulation of P in soil and water bodies. These phosphorus compounds, dissolved in surface or ground waters, are responsible for the eutrophication in closed water systems, especially in lakes and enclosed bays where the water is almost stagnant [4]. Phosphorus removal tech-
∗
Corresponding author. Tel.: +90 442 2314799; fax: +90 442 2360957. E-mail address: [email protected] (Y.S¸. Yildiz).
1383-5866/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2006.05.020
niques are chemical treatments like adsorption, chemical precipitation, ion exchange, electrodialysis, hybrid systems containing fly-ash adsorption and membrane filtration and electrocoagulation. Adsorption and chemical precipitation among the above methods have been widely used for phosphate removal [5–13]. The removal of phosphate from aqueous streams consists of the conversion of soluble phosphate to an insoluble solid phase. This solid phase can be separated from water by means of sedimentation or filtration. In wastewater applications, the most common and successful methods to precipitate phosphate involve the dissolved cations Al3+ , Ca2+ , Fe3+ and to a lesser extent of Fe2+ . It was found that when iron and aluminum are present in the water, FePO4 and AlPO4 forms in the low pH range (<6.5) and at higher pH range (>6.5) iron and aluminum increasingly convert to oxides and hydroxides. A higher pH is more ideal for precipitation of phosphate with calcium as apatites and hydroxyapatites [3]. In recent years, electrocoagulation has been successfully tested to treat wastewater. Electrocoagulation is a process consisting of creating metallic hydroxide flocks within the wastewater by electrodissolution of soluble anodes, usually made of iron or aluminum [14]. The difference between electro-
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
395
coagulation and chemical coagulation is mainly in the way of aluminum ions are delivered. In electrocoagulation, coagulation and precipitation are not conducted by delivering chemicals – called coagulants – to the system, but via electrodes in the reactor [14]. Electrocoagulation is based on the fact that the stability of colloids, suspensions and emulsions is influenced by electric charges. Therefore, if additional electrical charges are supplied to the charged particles via appropriate electrodes, the surface charge of particles is neutralized and several particles combine into larger and separable agglomerates [15]. Electrode assembly is the heart of the treatment facility. Therefore, the appropriate selection of its materials is very important. The most common electrode materials for electrocoagulation are aluminum and iron. They are cheap, readily available, and proven effective [16]. When aluminum is used as electrode material, the reactions are as follows: • At the cathode: 3H2 O + 3e− → 23 H2 (g) + 3OH−
Fig. 1. Schematic diagram of the experimental setup.
(1)
• At the anode: Al → Al3+ + 3e−
(2)
• In the solution: Al3+ (aq) + 3H2 O → Al(OH)3 + 3H+ (aq)
(3)
Taguchi’s orthogonal array (OA) analysis is used to obtain the best parameters for the optimum process design with the least number of experiments. In recent years, the Taguchi method has been used to determine optimum parameters because of its many advantages [17]. The main advantages of this method over other statistical experimental design methods are that the parameters affecting an experiment can be investigated as controlling and not controlling and that the method can be applied to an experimental design involving a large number of design factors [18,19]. Aim of this study is not to investigate the treatability of the phosphate containing wastewater by electrocoagulation method. Our aim is to determine the optimum operating conditions such as initial phosphate concentration, supporting electrolyte type and concentration, current density and initial pH of the wastewater for the removal of phosphate from waters by electrocoagulation method with plate aluminum electrodes based on removal efficiency.
filter with the pore diameter of 045 m (Schleicher and Schuell) before the analysis. The analysis of phosphate was carried out using the yellow vanadomolybdophosphoric acid method by a double beam spectrophotometer (Shimadzu UV-160 A) according to the Standard Methods for Examination of Water and Wastewater [20]. The initial pH was adjusted to a desired value using NaOH (Merck, 5N) or HNO3 (Carlo Erbaa, 65%). 2.2. Experimental setup and procedure
2. Experimental
The experimental setup is schematically shown in Fig. 1. The electrocoagulation unit consists of six pair of electrodes made of plate aluminum with total area of approximately 1500 cm2 and the gap between the electrodes is 5 mm. Electrodes were connected to a digital dc power supply (Shenzen-Mastech HY 3005-3) in monopolar mode. Two digital multimeters (Brymen Bm 201) as ampermeter and voltmeter were used to measure the current passing through the circuit and the applied potential, respectively. The electrocoagulation unit has been stirred at 150 rpm by a magnetic stirrer (Heidolp MR 3004 S). The experimental setup is shown in Fig. 1. The thermostated electrocoagulator is made of plexiglass with the volume of 850 mL. During the experiments, temperature, conductivity and pH of the wastewaters were measured by a multi-parameter (WTW Multiline P-4 F-Set-3). Reactor was operated in batch and galvanostatic mode.
2.1. Materials
2.3. Statistical analysis
All chemicals used were analytical grade and used without any further treatment. Distilled water was used in all experiments. Phosphate solutions were prepared from KH2 PO4 (Riedel de Ha¨en, 98%). NaCl (Merck, 99.5%), NaNO3 (Merck, 99%), Na2 SO4 (Sigma–Aldrich, 99%) and CaCl2 (Merck, >90) were used as supporting electrolyte. Treated wastewater was collected over a desired period of time from the reactor and collected samples were filtered by the cellulose acetate membrane
The variables chosen for this investigation are supporting electrolyte type and concentration, current density, initial phosphate concentration, and initial pH of the wastewater. The variables investigated and their levels were summarized in Table 1. Reaction period was kept constant in 25 min for statistical analysis. The use of the parameter design in the Taguchi method to optimize a process with multiple performance characteristics
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
396
Table 1 Variables and their values corresponding to their levels investigated in the experiments Variables
Levels
(A) Initial phosphate concentration (PO4 –P mg/L) (B) Initial pH of the wastewater (C) Supporting electrolyte concentration (mM) (D) Supporting electrolyte type (E) Current density (mA/cm2 )
1
2
3
4
50 4 0.0 NaCl 0.25
100 5 2.5 NaNO3 0.50
200 6 5.0 Na2 SO4 0.75
500 7 10.0 CaCl2 1.00
includes the following steps: (1) Identification of the performance characteristics and selection of the process parameters to be evaluated. (2) Determination of the number of parameter levels for the process and possible interaction between the process parameters. (3) Selection of the appropriate orthogonal array and assignment of process parameters to the orthogonal array. (4) Conduction of the experiments based on the arrangement of the orthogonal array. (5) Calculation of the performance characteristics. (6) Analysis of the experimental results by using the performance characteristics and ANOVA. (7) Selection of the optimal levels of process parameters. (8) Verification of the optimal process parameters through the confirmation experiment [21,18]. The orthogonal array (OA) experimental design was chosen as the most suitable method to determine an experimental plan, L16 (45 ) (Table 2), for five parameters each with four values [21]. The experimental variables, their levels and results of conducted experiments are given in Table 2. In order to observe the effects of noise sources on the electrocoagulation process, each experiment was repeated twice under
the same conditions at different times. The performance characteristics were chosen as the optimization criteria. There are three categories of performance characteristics: the larger the better – this category was used to evaluate the system performance based on phosphate removal efficiency – the smaller the better and the nominal the better. The two performance characteristics were evaluated by using Eqs. (4) and (5) [21]: • Larger is better
SNL = −10 log10
n
1 1 n Yi2
(4)
i=1
• Smaller is better SNS = −10 log10
n
1 2 Yi n
(5)
i=1
where SNL and SNS are the performance characteristics, n the number of repetitions done for an experimental combination and Yi is the performance value of the ith experiment. In the Taguchi method, the experiment corresponding to optimum working conditions might not have been done during the whole period of the experimental stage. In such cases, the performance value corresponding to the optimum working conditions can be predicted by utilizing the balanced characteristic of OA. For
Table 2 Experimental variables, their levels and results of conducted experiments corresponding to L16 experimental plan Experiment no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Variables and their levels
Removal efficiency of phosphate (%)
(A)
(B)
(C)
(D)
(E)
First series
Second series
Average
1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1 2 3 4 2 1 4 3 3 4 1 2 4 3 2 1
1 2 3 4 3 4 1 2 4 3 2 1 2 1 4 3
1 2 3 4 4 3 2 1 2 1 4 3 3 4 1 2
99.3 100.0 96.9 100.0 97.2 97.0 95.2 78.0 100.0 48.0 86.6 71.0 64.7 73.0 37.0 26.2
93.3 100.0 98.6 100.0 95.6 97.8 95.7 72.0 99.7 63.0 85.5 67.0 62.8 72.0 26.3 22.3
96.3 100.0 97.7 100.0 96.4 97.4 95.5 75.0 99.9 55.7 86.1 69.0 63.7 72.5 31.7 24.2
Energy consumption (kW h/m3 ), average 2.86 1.91 22.07 2.27 7.03 20.96 0.89 0.80 0.83 5.56 11.42 2.60 3.20 4.12 0.41 19.89
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
397
Table 3 Results of the analysis of variance for the removal efficiencies of phosphate Variables
Sum of squares
Degrees of freedom
Mean of squares
F
(A) Initial phosphate concentration (PO4 –P mg/L) (B) Initial pH of the wastewater (–) (C) Supporting electrolyte concentration (mM) (D) Supporting electrolyte type (–) (E) Current density (mA/cm2 ) Error
11895.6 2008.3 674.5 1156.7 2489.8 228.2
3 3 3 3 3 16
3965.2 669.4 224.8 385.6 829.9 14.3
278.08 46.95 15.77 27.09 58.20
this, the following additive model may be used: Yi = m + X i + e i
(6)
where m is the overall mean of performance value, Xi the fixed effect of the parameter level combination used in the ith experiment and ei is the random error in the ith experiment. If experimental results are stated in a percentage (%), before evaluating Eq. (6), the Ω transformation of percentage values should be applied first using Eq. (7) by which values of interest are also determined later by carrying out a reverse transformation by using the same equation [22]: 1 Ω(db) = −10 log −1 (7) P where Ω(db) is the decibel value of percentage value subject to omega transformation and P is the percentage of the product obtained experimentally. Since Eq. (6) is a point estimation which is calculated by using experimental data in order to determine whether the additive model is adequate or not, the confidence limits for the prediction error must be evaluated [21]. The prediction error is the difference between the observed Yi and the predicted Yi . The confidence limits for the prediction error are 1 1 2 σ + σ2 Se = ±2 (8) n0 e nr e sum of squares due to error (9) degrees of freedom for error 1 1 1 1 1 1 1 1 + + + ··· = + − − − n0 n nAi n nBi n nCi n (10)
σe2 =
where Se is the two-standard deviation confidence limit, n the number of rows in the matrix experiment, nr the number of repetitions in the confirmation experiment and nAi , nBi , nCi , . . . are the replication numbers for the parameter levels Ai, Bi, Ci, . . . If the prediction error is outside these limits, the possibility that the additive model is not adequate should be suspected. Otherwise, the additive model can be considered to be adequate. A verification experiment is a powerful tool for detecting the presence of interactions among the control parameters. If the predicted response under the optimum conditions does not match the observed response, it implies that the interactions are important. If the predicted response matches the observed response, it then implies that the interactions are probably not important and that the additive model is a good approximation [29]. The order of the experiments was obtained by inserting
parameters into the columns of OA and L16 (45 ) which were chosen as the experimental plan given in Table 2. The order of experiments was made random in order to avoid noise sources which had not been considered initially and which could take place during an experiment and affect results in a negative way. 3. Results and discussions 3.1. Statistical analysis The collected data were analyzed by an IBM compatible PC. In order to see effective parameters and their confidence levels on the electrocoagulation process, the analysis of variance was performed. A statistical analysis of variance (ANOVA) was performed to see whether the process parameters were statistically significant or not. The F-test is a tool to see which process parameters have a significant effect on the removal efficiency. The F-value for each process parameter is simply a ratio of mean of the squared deviations to the mean of squared error. Usually, the larger the F-value, the greater the effect on the performance criteria value due to the change of the process parameter. With the performance characteristics and ANOVA analyses, the optimal combination of process parameters can be predicted [17]. The results of variance analysis are given in Table 3. F values in the table show that the effective parameters on the removal of phosphate by electrocoagulation method are initial phosphate concentration, current density, initial pH of the wastewater and supporting electrolyte type. To obtain optimal phosphate removal performance, the larger the better performance characteristic (Eq. (4)) has been taken for the removal of phosphate. The optimal level of a process parameter is the level with the highest SN value calculated by Eq. (4). Fig. 2 shows the variation of the performance characteristics with the variables. To determine the experimental conditions for the first data point, the CD
Fig. 2. The effect of each parameter on the optimization criteria for phosphate removal.
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
398
Table 4 Optimum working conditions and alternative working conditions for different experimental setups observed and predicted removal efficiencies of phosphate Case no.
1* 2 3 4 5 6 7 8 9 10 11 *
A (Cf )
B (pHi )
OBS.
PRED
CONF
Level
Value
Level
Value
C (sec) Level
Value
D (set) Level
Value
E (CD) Level
Value
–
–
–
1 1 1 1 1 2 1 2 2 2 3
50 50 50 50 50 100 50 100 100 100 200
1 1 1 1 1 1 1 4 3 1 1
4 4 4 4 4 4 4 7 6 4 4
3 4 3 3 4 2 4 3 2 2 2
5 10 5 5 10 2.5 10 5 2.5 2.5 2.5
1 4 4 2 2 3 1 1 4 1 1
NaCl CaCl2 CaCl2 NaNO3 NaNO3 Na2 SO4 NaCl NaCl CaCl2 NaCl NaCl
4 4 4 4 4 4 1 2 1 1 1
1 1 1 1 1 1 0.25 0.50 0.25 0.25 0.25
100.0 100.0 100.0 98.0 98.2 96.9 93.9 88.0 99.5 87.0 63.0
99.9 100.0 100.0 100.0 100.0 96.4 99.4 84.5 76.0 87.3 76.8
88.9–100.0 89.0–100.0 89.0–100.0 89.0–100.0 89.0–100.0 85.4–100.0 88.34–100.0 73.5–100.0 65.0–87.5 76.3–98.4 65.8–87.8
Optimum working conditions
(current density) for that point is level 1 which is 0.25 mA/cm2 for this parameter. The experiments for which CD level is 1 are experiments 1, 8, 10 and 15. The performance characteristics value of the first data point is thus the average of those obtained from experiments 1, 8, 10 and 15. Thus, experimental conditions for the fourth data point are the conditions of the experiments for which column CD is 2 (experiments 4, 5, 11 and 14) [23,24]. The numerical value of the maximum point in each graph marked the best value of that particular parameter and was found as A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and E4(1 mA/cm2 ). These parameter values provide the optimum conditions. If experimental plan given in Table 2 is studied carefully together with parameter values given A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and CD4(1 mA/cm2 ) it can be seen that experiments corresponding to optimum conditions A1(50 PO4 –P mg/L), B1(4), C3(5 mM), D1(NaCl) and CD4(1 mA/cm2 ) have not been carried out during the experimental work. Thus, it should be noted that the removal efficiencies for phosphate (%) in Table 4 are predicted results obtained by using Eq. (6) and observed results for the same conditions. Also, the results in Table 4 are confidence limits of predictions. In order to test the predicted results, confirmation experiments were carried out once at the same working conditions. The fact that the removal efficiencies from the confirmation experiments are within the calculated confidence intervals calculated from Eqs. (8)–(10) (Table 4) shows that the experimental results are within ±5% in error. This case states that there is a good agreement between the predicted values and experimental values and the interactive effects between the parameters are indeed negligible. It may be concluded that the additive model is adequate for describing the dependence of the dissolution process on the various parameters [21].
the preceding equation can be integrated to give the following: C0 = K1 At (12) V ln Ct It can be concluded that Eq. (12) can be applied to removal of phosphate from waters by electrocoagulation. Fig. 3 illustrates the variation of removal efficiency (η) with time. Figs. 3 and 4 were plotted using the data obtained from experiment 3 and similar tendencies were observed in other test runs (not shown). Sharp increases of removal efficiencies are clearly observed initially. After the treatment period of 15 min the removal efficiencies approach plateaus at 90% (the point with arrow in Fig. 3). The slight improvement for phosphate removal following a great increase in the energy consumption (see Fig. 4) and is not worth the energy consumed because of the rapid increases in applied potential as shown in Fig. 4. Removal efficiencies and energy consumptions increases with increasing time, but increase in energy consumption is higher than that of removal efficiency beyond the plateau in Fig. 3. For example, system consumes electrical energy of 14.37 kW h/m3 and provides removal efficiency of 94.4% for 16 min while energy consumptions raise from 14.37 kW h/m3
3.2. Kinetic analysis of the electrocoagulation process According to Hosny, the rate of oil removal from oil–water emulsions can be expressed by Eq. (11) [25]: dC −V (11) = K1 AC dt
Fig. 3. Variation of phosphate removal efficiency with time (C0 = 50 mg/L, pHi = 6.0, CSE = 5.0 mM, Na2 SO4 , CD = 0.75 mA/cm2 ).
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
399
Table 5 Variation of removal rate constants with experimental variables corresponding to L16 experimental plan Experiment no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Variables and their levels (A)
(B)
(C)
(D)
(E)
1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
1 2 3 4 2 1 4 3 3 4 1 2 4 3 2 1
1 2 3 4 3 4 1 2 4 3 2 1 2 1 4 3
1 2 3 4 4 3 2 1 2 1 4 3 3 4 1 2
η (%), average
t (min), average
K1 (cm/min), average
K1 /I (×10,000 cm/A-min), average
94.4 85.0 85.5 91.8 96.4 93.4 92.7 77.1 98.1 76.5 93.4 87.4 77.6 89.2 67.8 54.6
22 12 19 4 25 18 23 40 20 50 32 40 50 55 80 60
886.3 976.9 678.6 3935.8 733.8 883.7 663.0 227.8 916.0 171.7 480.5 277.4 181.3 236.9 74.8 83.0
2363.4 1302.6 603.2 2623.9 489.2 785.5 884.0 607.5 1221.4 457.9 320.4 246.6 161.2 157.9 199.5 110.7
All calculations conducted for K1 values were performed based on the time for reach the beginning of the plateau and presented in Table 5. When examined Table 5, it can be seen that cell removal constant has been strongly affected by the current passing through the circuit. Hence, it is more favorable that Eq. (12) should be modified as noted Eq. (13): V ln
Fig. 4. Variation of specific energy consumption with time (C0 = 50 mg/L, pHi = 6.0, CSE = 5.0 mM, Na2 SO4 , CD = 0.75 mA/cm2 ).
to about 22 kW h/m3 removal efficiencies increase from 94.4% to only 98.6% between the 16 and 25 min. Thus the system should be operated at the end of the sharp increase step.
C0 Ct
= K1 AIt
(13)
Maximum values of cell removal constants are achievable, when CaCl2 (Merck, 90%) is used as supporting electrolyte (see Table 5, exp. nos. 4, 9 and 15). Cell removal constants calculated according to optimum and alternative working conditions have been presented in Table 6. It can be said that the tendency based on cell removal constants in Table 6 is the same as Table 5 (see Table 6, exp. nos. 2, 3 and 9). Cell removal constants calculated for CaCl2 have the maximum values within the groups arranged according to initial phosphate concentrations. For example, in the initial phosphate
Table 6 Cell removal constants calculated according to optimum and alternative working conditions Experiment no.
1* 2 3 4 5 6 7 8 9 10 11 *
Variables and their levels (A)
(B)
(C)
(D)
(E)
1 1 1 1 1 2 1 2 2 2 3
1 1 1 1 1 1 1 4 3 1 1
3 4 3 3 4 2 4 3 2 2 2
1 4 4 2 2 3 1 1 4 1 1
4 4 4 4 4 4 1 2 1 1 1
Optimum working conditions
η (%)
t (min)
96.6 97.9 97.3 91.2 91.1 96.0 89.6 85.0 96.2 89.8 83.9
10 4 4 10 6 8 16 20 16 28 50
K1 /I (×10,000 cm/A-min)
4630.5 12942.0 12814.5 3567.0 6526.5 6078.0 7710.0 3006.0 26106.0 24312.0 2220.0
400
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
Fig. 6. X-ray diffractogram of precipitate obtained from the experiments (100 mg/L PO4 –P).
Fig. 5. (a) Variation of wastewater pH with time (experiments 1, 5 and 6 are confirmation test runs presented in Table 6); (b) variation of wastewater pH with time (experiments 2, 3 and 9 are confirmation test runs presented in Table 6).
concentration of 50 mg/L; cell removal constants for exp. nos. 2 and 3 equal to 12,942 and 12814.5 cm/A-min, respectively. Similarly 100 mg/L of initial phosphate concentration; cell removal constant is 26,106 cm/A-min for exp. no. 9. Hydroxyapatites, which are slightly soluble, form [26] according to reaction (14) [27] within the pH range of 4.0–8.5. When CaCl2 is used as supporting electrolyte: 5Ca2+ + 7OH− + 3H2 PO4 − → CaOH(PO4 )3 + 6H2 O
(14)
Variation of the wastewater pH with time has been presented in Fig. 5a. In all test runs which have been used as supporting electrolyte different from CaCl2 , waste water pH gradually increases with time. For example, within 20 min, pH values of the wastewater have increased approximately unit of 4.35, 4.60 and 4.00 for the experiment nos. 1, 5 and 6, respectively. Whereas wastewater pH is slightly increases with time when CaCl2 is used In the first 20 min, the pH values of the wastewater have increased unit of about 1, 2 and 1.2 for the experiment nos. 2, 3 and 9, respectively (see Fig. 5b). This behavior of the system can be attributed to the reaction (14) as seen in reaction (14), in order to form 1 mol of hydroxyapatite, 7 mol of OH− ions should be consumed, hence pH of the wastewater is not be able to increase. In order to determine the species in the precipitate, X-ray diffractograms have been used. The results of the X-ray
diffractogram of the precipitate obtained from the experiments conducted with containing100 mg/L PO4 –P have been given in Fig. 6. As seen in Fig. 6, Al(OH)3 and AlPO4 are available in the medium. The precipitate produced by the process of phosphate removal using electrocoagulation method is usually a mixture of Al(OH)3 and AlPO4 although the AlPO4 precipitation is favored over Al(OH)3 [27], when other ions are absence in the wastewater namely not using any supporting electrolyte. Preliminary investigations carried out by the authors have shown that aluminum electrodes provided the better results comparing to iron electrodes. Aluminum electrodes have higher removal efficiency and lower energy consumption than those of iron electrodes [28,29] and The effluent with aluminum electrodes was found very clear and stable, whereas the effluent with iron electrodes appeared greenish first, and then turned yellow and turbid. The green and yellow colors must have resulted from Fe(II) and Fe(III). Fe(II) is the common ion generated in situ of electrolysis of iron [16]. 4. Conclusions In this investigation, the Taguchi method was used to determine the optimum conditions for the phosphate removal. Effect of initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density on the electrocoagulation of phosphate ion has been investigated and effects of these parameters on the system performance have been evaluated based on removal efficiency. It can be said that the Taguchi method is able to use for the optimization of phosphate ions from wastewaters by electrocoagulation method: 1. The effective parameters on the removal of phosphate from wastewaters by electrocoagulation are initial phosphate concentration, current density and initial pH of the solution. 2. The optimum conditions for the parameters initial phosphate concentration, initial pH of the wastewater, supporting electrolyte concentration, supporting electrolyte type and current density were equal to 50 mg/L, 4, 5 mM NaCl and 1 mA/cm2 ,
˙ S¸. Irdemez et al. / Separation and Purification Technology 52 (2006) 394–401
respectively. Under these conditions, a removal efficiency of 100% can be achieved (Table 4). 3. It can be said that maximum cell removal constant values are achievable in the experiment groups arranged according to initial phosphate concentrations when CaCl2 is used as supporting electrolyte. Besides electrocoagulation system can reach the plateau in the shortest treatment periods (see Table 5), so detailed cost analysis of the system should be conducted and using potential of the CaCl2 as supporting electrolyte should be evaluated. 4. It can be concluded that the phosphate can be removed completely from water by using electrocoagulation. In electrocoagulation, the main reaction is phosphate removal has been accomplished accompanying with the precipitation of Al(OH)3 and AlPO4 although the AlPO4 precipitation is favored over Al(OH)3 and detailed economic analysis of the whole process is necessary for a more precise evaluation of the process (see Table 2 for energy consumptions). 5. The predicted and observed removal efficiency values are close to each other. Thus, it may be concluded that the additive model is adequate for describing the dependence of the removal efficiencies on the various parameters. Acknowledgements Authors are grateful to the research council of Atat¨urk University, for providing financial support with grant no. 2002/143. The authors also appreciate Dr. Ferhat B¨ulb¨ul for providing the valuable comments for evaluation of X-ray diffractograms. References [1] C. Sommariva, A. Converti, M. Del Borghi, Increase in phosphate removal from wastewater by alternating aerobic and anaerobic conditions, Desalination 108 (1996) 255–260. [2] G. Tchobanoglous, F.L. Burton, Wastewater Engineering, McGraw-Hill, 1991. [3] D.G. Grubb, M.S. Guimaraes, R. Valencia, Phosphate immobilization using an acidic type F fly ash, J. Hazard. Mater. 76 (2000) 217– 236. [4] S.H. Lee, W.H. Moon, Adsorption of phosphorus in saturated slag media columns, Sep. Purif. Technol. 12 (1997) 109–118. [5] L. Johansson, J.P. Gustafsson, Phosphate removal using blast furnace slags and opoka-mechanisms, Water Res. 34 (2000) 259. [6] N.M. Agyei, C.A. Strydom, J.H. Potgieter, An investigation of phosphate ion adsorption from aqueous solution by fly ash and slag, Cem. Concr. Res. 30 (2000) 823. [7] M. Rao, A.V. Parwate, A.G. Bhole, Removal of Cr6+ and Ni2+ from aqueous solution using bagasse and fly ash, Waste Manage. 22 (2002) 821–830.
401
[8] C.J. Lin, J.E. Chang, Effect of fly ash characteristics on the removal of Cu(II) from aqueous solution, Chemosphere 44 (5) (2001) 1185–1192. [9] B. Bayat, Combined removal of zinc(II) and cadmium(II) from aqueous solutions by adsorption onto high-calcium Turkish fly ash, Water Air Soil Pollut. 136 (2002) 69–92. [10] P.C. Kao, J.H. Tzeng, T.L. Huang, Removal of chlorophenols from aqueous solution by fly ash, J. Hazard. Mater. 76 (2000) 237–249. [11] S¸. ˙Irdemez, Y.S¸. Yıldız, N. Demircio˘glu, Phosphate removal from wastewaters by using fly-ash adsorption, in: Proceedings of the International Soil Congress, Erzurum, Turkey, 2004, pp. 29–34. [12] E. Yıldız, Phosphate removal from water by fly ash using crossflow microfiltration, Sep. Purif. Technol. 35 (2004) 241–252. [13] S¸. ˙Irdemez, Phosphate removal from wastewaters by electrocoagulation, Ph.D. Thesis, School of Natural and Applied Science, Atat¨urk University, Erzurum, 2005. [14] A.S. Koparal, The removal of salinity from produced formation water by conventional and electrochemical methods, Fresen. Environ. Bull. 11 (12a) (2002) 1071–1077. ¨ gu¨ tveren, S. Koparal, Electrocoagulation for oil-water em¨ulsion ¨ [15] U.B. O˘ treatment, J. Environ. Sci. Health A 32 (9/10) (1997) 2507–2520. [16] X. Chen, G.C. Chen, P.L. Yue, Separation of pollutants from restaurant wastewater by electrocoagulation, Sep. Purif. Technol. 19 (2000) 65–76. [17] C.Y. Nian, W.H. Yang, Y.S.J. Tarng, Optimization of turning operations with multiple performance characteristics, Mater. Process. Technol. 95 (1999) 90. ¨ K¨uc¸u¨ k, M.M. Kocakerim, M. C [18] O. ¸ opur, A. Yartas¸ı, Optimization of dissolution of ulexite in (NH4 )2 SO4 solutions, Can. Metall. Quart. 44 (2005) 1–6. ¨ K¨uc¸u¨ k, Application of Taguchi method in the optimization of dissolution [19] O. of ulexite in NH4 Cl solutions, Korean J. Chem. Eng. 23 (1) (2006) 21–27. [20] Standard Methods for Water and Wastewater Examination, 16th ed., AWWA, APHA, WPCF, New York, 1985. [21] M.S. Phadke, Quality Engineering Using Robust Design, Prentice Hall, New Jersey, 1989, pp. 61–292. [22] G. Taguchi, System of Experimental Design, Quality Resources, New York, 1987, p. 108. [23] M. C ¸ opur, T. Pekdemir, C. C ¸ elik, S. C ¸ olak, Determination of the optimum conditions for the dissolution of stibnite in HCl solutions, Ind. Eng. Chem. Res. 36 (3) (1997) 682–687. [24] M. C ¸ opur, An optimization study of dissolution of Zn and Cu in ZnS concentrate with HNO3 solutions, Chem. Biochem. Eng. 15 (4) (2002) 191–197. [25] A.Y. Hosny, Separating oil from oil-water emulsions by electroflotation technique, Sep. Technol. 6 (1996) 9–17. [26] W.J. Eilbeck, G. Mattock, Chemical Processes in Wastewater Treatment, John Wiley & Sons, New York, 1987. [27] W.W. Eckenfelder Jr., Industrial Water Pollution Control, 2nd ed., McGrawHill, 1989. [28] S¸. ˙Irdemez, N. Demircio˘glu, Y.S¸. Yıldız, The effects of pH on phosphate removal from wastewater by electrocoagulation with iron plate electrodes. J. Hazard. Mater., 2006, in press. [29] S¸. ˙Irdemez, N. Demircio˘glu, Y.S¸. Yıldız, Z. Bing¨ul, The effects of current density and phosphate concentration on phosphate removal from wastewater by electrocoagulation using aluminum and iron plate electrodes, Sep. Purif. Technol. 52 (2006) 218–223.