CC BY-NC-ND
0ELSEVIER
CrossMark
Available online at www.sciencedirect.com
ScienceDirect
Karbala International Journal of Modern Science 1 (2015) 78—85
http://www.journals.elsevier.com/karbala-international-journal-of-modern-science/
Evaluating the TiO2 as a solar photocatalyst process by response surface methodology to treat the petroleum waste water
Dheeaa al deen Atallah Aljuboury a *, Puganeshwary Palaniandy a, Hamidi Bin Abdul Aziz a, Shaik Feroz b
a School of Civil Engineering, Universiti Sains Malaysia, Malaysia b Caledonian College of Engineering, Oman
Received 21 July 2015; revised 7 October 2015; accepted 22 October 2015 Available online 28 November 2015
Abstract
The aim of this study is to investigate the performance of employing the solar photo-catalyst of TiO2 to treat petroleum wastewater from Sohar oil Refinery (SOR), evaluate the performance of employing this process by a central composite design (CCD) with response surface methodology (RSM) and evaluate the relationships among operating variables such as TiO2 dosage, pH, C0 of COD, and reaction time to identify the optimum operating conditions. Quadratic models prove to be significant with very low probabilities (<0.0001) for the following two responses: total organic carbon (TOC) and chemical oxygen demand (COD).
TiO2 dosage and pH are the two main factors that improved the TOC and COD removal while C0 of COD and reaction time are the actual factors. The optimum conditions are a TiO2 dosage (0.6 g/L), C0 of COD (1600 ppm), pH (8), reaction time (139 min) in this method. TOC and COD removal rates are 15.5% and 48.5%, respectively. The predictions correspond well with experimental results (TOC and COD removal rates of 16.5%, and 45%, respectively). Using renewable solar energy and treating with minimum TiO2 input make this method to be a unique treatment process for petroleum wastewater.
© 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of University of Kerbala. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Petroleum waste water; Photo-catalyst of TiO2; Advanced oxidation process (AOP); Response surface methodology (RSM)
1. Introduction
Nowadays, one of the major problems facing industrialized nations is contamination of the environment by hazardous chemicals. A wide range of pollutants compounds are detected in petroleum waste
* Corresponding author. School of Civil Engineering, Universiti Sains Malaysia, Nibong Tebal, Seberang Perai Selatan, Pulau Pinang, 14300, Malaysia. Tel. +60173498238; Oman. Tel. +96895358288. E-mail address: msc.dheeaa@yahoo.com (D.A. Aljuboury). Peer review under responsibility of University of Kerbala.
water in Sohar Oil Refinery, so, the elimination of these chemicals from petroleum wastewater is presently one of the most important aspects of pollution control in Oman.
Advanced oxidation processes (AOPs) have capability of rapid degradation of recalcitrant pollutants in the aquatic environment. Remediation of hazardous substances is attributed to hydroxyl radical (*OH), which exhibits reactivity toward organic. Many technical enhanced the production rate of *OH by chemical additives (such as H2O2), external energy (such as UV and sunlight), catalysts (such as TiO2) and the
http://dx.doi.org/10.1016/j.kijoms.2015.10.006
2405-609X/© 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of University of Kerbala. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
integration of two or more AOPs (such as TiO2/Fenton/ sunlight) [1].
Photo-catalysis is a promising technique for the treatment of contaminated water, which has been widely studied in recent years because it is fast, effective, eco-friendly, economically viable and able to completely oxidize organic molecules at a low energy cost [2].
Several previous studies have reported the enhanced oxidation of contaminants by photo-catalyst of TiO2. It has been found that solar photo-catalytic oxidation process was effective in treating a synthetic high COD wastewater, it effectively reduced the COD content by 86% [3]. Javad Saien and Fatemeh Shahrezaei [8] showed that 78% of the COD was removed from real refinery wastewater, containing a range of aliphatic and aromatic organic compounds which were collected from the Kermanshah (Iran) refinery plant, by using the photo-catalyst in UV/TiO2 process. Jose [4] reported that the use of solar photocatalysis in the presence of TiO2 constituted a very effective and rapid method for the reduction and even elimination of these pesticides in leaching water. Ramanjot [5] showed that 83% of the Procion Blue (PB) dye was removed from industrial wastewater, containing a range of dyes which were collected from the collected from textile mill (Punjab), by using the photo-catalyst in sunlight/TiO2 process. Santos [6] reported that photocatalysis (TiO2/ UV) achieved High rates of removal for phenols, oil and grease, and dissolved organic carbon from petroleum refinery wastewaters which were collected from the a Brazilian oil refinery plant. Shanmuga [7] carried out solar photo-catalytic experiments with 0.2 g/l of TiO2 catalyst for different concentrations of phenol wastewater and it was found that complete degradation of phenol was possible in a reasonable time (less than 300 min) when concentration of phenol was <100 ppm.
This work evaluates the solar photo-catalyst of TiO2 on the degradation of COD and TOC in petroleum waste water. So, the main aims for this study are as follows:
• To evaluate the performance of employing the solar photo-catalyst of TiO2 by a central composite design (CCD) with response surface methodology (RSM) to degradation of TOC and COD from the petroleum wastewater.
• To evaluate the statistical relationships among operating variables (such as TiO2 dosage, pH, C0 of COD, and reaction time) and the responses, which COD & TOC removal efficiencies are selected as the responses for optimization.
• To determine the optimum operational conditions of this method.
2. Materials and methods
2.1. Wastewater characterization
The physicochemical characteristics of the petroleum wastewater from Sohar oil refinery (SOR) are summarized in Table 1. The samples of raw effluent are collected in different days from the point that the wastewater is just leaving the dissolved air flotation (DAF) and just into the biological treatment unit in wastewater treatment plant at Sohar petroleum refinery. Samples are transferred to the laboratory and stored under refrigeration (4 °C) until use. Samples are characterized before the experiments to obtain their chemical and physical properties. Petroleum waste water characterization is determined by the quantification of pH, chemical oxygen demand (COD), and Total Organic Carbon (TOC) according to the Standard Methods for the Examination of Waste water methodology.
3. Materials
The catalyst is TiO2 Aeroxide P-25 (manufactured by Evonik Industries Co in Germany). They are used for the solar photo-catalyst of TiO2 process to degradation of TOC & COD. Sulfuric acid and sodium hydroxide are used to adjust the pH to the desired values.
Table 1
Characteristics of petroleum wastewater from Sohar oil refinery (SOR).
No Parameter Range of concentrations in petroleum wastewater Average The standard discharge limit
1 pH 6—8 7 6—9
2 Conductivity (Micro S/cm) 2600—3950 3275 2000—2700
3 TDS (ppm) 1200—1500 1350 1500—2000
4 TOC (ppm) 220—265 243 50—75
5 COD (ppm) 550—1600 1075 150—200
3.1. Analytical procedure and characterization of the industrial effluent
A Shimadzu TOC analyzer (LCSH/CSN) is used to measure the TOC for each sample. Chemical oxygen demand (COD) is measured by COD photometer (manufactured by CHEMetrics). COD is estimated before and after treatment. Before each analysis, samples are filtered by filter papers (0.22 um Millipore Durapore membrane, 40 Ashless, Diameter 150 mm). Solar ultraviolet radiation (UV) is measured by a global UV radiometer (KIPP & ZONEN).
3.2. Experimental procedure
A sketch of the solar photo-catalyst of TiO2 process is shown in Fig. 1. It consists of a glass recirculation tank (1.5 L), which is subjected to stirring to maintain a well-mixed solution during the experiments, connects
to the tubular solar reactor (four tubes (50 cm length x 2 cm inner diameter x 0.1 cm thickness)). The solution is re-circulated through the reactor at a flow rate of about 1.5 L/min by means of a peristaltic pump. The catalyst materials are added in a glass recirculation tank during this process. The solar radiation intensity is approximately 650 W/m2 during experimental runs. All experimental are carried out in same duration (12 P.M. — 3 P.M.). The tubular photo reactor operate at a UV-index from 8 to 11 according to Exposure category is very high.
The UV-Index is calculated as follow:
Take the output from the UV-E radiometer according to ISO 17166:1999/CIE S007/E-1998. Transform the output voltage to W/m2 with the instruments sensitivity. Eq. (1) allows calculating the amount of UV intensity received on any surface in the same position with regard to the sun by UV-Index (UVI): UVI = R(W/m2)*40(m2/W) (1)
Fig. 1. A sketch of the solar photo-catalyst of TiO2 process.
1 2 3 4 5 6 7 8 9 10 11 12
| Low Moderate 1 High | Very high Extreme
Fig. 2. UV-index which measures UV intensity levels on a scale of 1—12.
where:
UVI is the UV-Index as shown in Fig. 2. R is the reading (R) in UV radiometer by (W/m2) unit.
The pH for petroleum wastewater samples is used between 6 and 9. Several set of experiments are carried out according to a central composite design (CCD) with response surface methodology (RSM) to determine the COD and TOC removal efficiency under the optimum operational conditions.
4. Results and discussion
4.1. Experimental design and the analysis of variance (ANOVA)
Central composite design and response surface methodology are employed in the statistical design of the experiments, data analysis, explaining the optimal conditions of the independent variables and assessment of the relationships among four significant independent variables, which are TiO2 dosage, pH, C0 of COD, and reaction time as shown in Table 2.
Each independent variable is varied over three levels according to face centered CCD as — 1, 0, and + 1, respectively at the determined ranges base on a set of preliminary experiments. The total number of experiments conducts for the four factors according to Eq. (2)
No. of Experiments = 2k + 2k + 6 (2)
where;
k is the number of factors.
Table 2
Central Composite Design (CCD) independent variables.
The factors Level of value
-1 0 + 1
pH 6 7.5 9
TiO2 (g/l) 0.5 1 1.5
COD (ppm) 850 1225 1600
RT (min) 60 120 180
a Reaction time.
The design consisted of 2k factorial points augmented by 2k axial points and 6 replications for a center point. In this work, the total number of experiments conducted for the four factors is 30 with 16 factorial points, 8 axial points and 6 replications to assess the pure error and get a good estimate. The COD and TOC removal are the dependent variables (responses) during this process. Performance is evaluated by analyzing the COD and TOC removal efficiencies as shown in Table 3.
The behavior of the system is explained through an empirical second-order polynomial model, as shown in Eq. (3) [10].
y=b0+E bjX+E biX+EE biiXiXi+e
j=1 j=1 i < j=2
where;
Y is the response.
Xi and Xj are the variables.
b is the regression coefficient.
k is the number of factors studied and optimize in
the experiment.
ei is the random error.
A total of 30 runs are executed using the Central Composite Design (CCD) experimental design; interactions among the four independent variables are considered in each run to investigate the validity of treating petroleum waste water using solar photo-catalyst of TiO2 during advanced oxidation. As shown in Table 3, the removal efficiencies range from 6.1% to 15.6% for TOC and 12% to 55.8% for COD.
The analysis of variance (ANOVA) is used for graphical analysis of data to obtain the interaction among the process variables and the responses. The quality of the fit polynomial model is expressed by coefficient of determination (R2). Model terms are evaluated by the P-value (probability) with 95% confidence level. The analysis of variance (ANOVA) for TOC and COD removal are represented in Table 4. All of the response surface quadratic models for parameters in this table are significant at the 5% confidence level since the P-values are less than 0.05. The
Table 3
Responses values for different experiment conditions.
Run Factors Responses
A B C D 1 2
TiO2 (g/l) pH COD (ppm) RT (min) Degradation of TOC, (%) Degradation of COD, (%)
1 0.5 6 1600 180
2 1 7.5 1225 120
3 1.5 7.5 1225 120
4 0.5 9 850 60
5 0.5 6 850 60
6 1.5 6 1600 60
7 1.5 6 1600 180
8 0.5 7.5 1225 120
9 0.5 9 1600 60
10 1 7.5 850 120
11 1 7.5 1600 120
12 1.5 9 850 180
13 1.5 9 850 60
14 1 7.5 1225 120
15 1 6 1225 120
16 0.5 6 1600 60
17 1.5 9 1600 180
18 1.5 6 850 180
19 1.5 9 1600 60
20 1 7.5 1225 120
21 0.5 6 850 180
22 1 7.5 1225 120
23 1 9 1225 120
24 1.5 6 850 60
25 0.5 9 850 180
26 1 7.5 1225 120
27 1 7.5 1225 60
28 1 7.5 1225 120
29 0.5 9 1600 180
30 1 7.5 1225 180
correlation coefficients (R2) for the TOC and COD
removal rates are 0.995, and 0.869, respectively, which they are greater than 0.80, the cut-off for a model with good fit. A high coefficient (R2) value ensures a satisfactory adjustment of the quadratic model to the experimental data and illustrates good agreement between the calculated and observed results and shows that a desirable and reasonable agreement with the adjusted R2 [9,11,12]. If the model terms have the P-value (probability) more than 0.05, they are considered limited influence. So, they must be excluded from the study to improve the models. The model of TOC and COD removal are considered significant using the F-test at 5% significant level (Prob < 0.05). The "adequate precision" (AP) ratio values which are higher than 4 are desirable and confirm that the predicted models can be used to navigate the space defined by the central composite design (CCD) [9]. The "adequate precision" (AP) ratio of the models in this study is adequate, which vary between 92.71 and 13.42.
14.93 55.2
15.2 25.5
9.1 21
6.1 27
10.5 33
13.5 40.5
15.6 49.5
10.7 33
9.3 12
14.95 55.5
10.4 39
8.5 25.5
12.7 33
9.9 24
11.3 24
14.9 55.65
7.5 21
8.5 15
10.5 42
15.2 55.8
12.4 30
15.1 55.35
12.5 48
In the current study, two quadratic models are significant model terms (Table 4). Insignificant model terms, which have limited influence, are excluded from the study to improve the models. Based on the results obtained, the response surface models for predicting TOC and COD removal efficiencies are considered reasonable.
4.2. Normal probability
The final regression models are presented in terms of their coded and actual factors (Table 5). Normal probability plots of the studentized residuals and diagnostics are provided by Design Expert 6.0.7 software program (a statistical software package from Stat-Ease Inc) to confirm that the selected models provide an adequate approximation of the real system. Plots of Normal probability aid in evaluating the models. Fig. 3 demonstrates the normal probability plots for the studentized residuals for TOC and COD removal efficiencies. The normal probability plots
Table 4
ANOVA results and adequacy of the quadratic models for TOC and COD removal efficiencies.
Source Sum of squares Degrees of freedom Mean square F-value Prob > F
Model 201.6712 7 28.81017 712.4886 <0.0001
A 6.600556 1 6.600556 163.2347 <0.0001 Mean 11.90933
B 13.69389 1 13.69389 338.6561 <0.0001 R2 0.995608
C 20.05556 1 20.05556 495.983 <0.0001 Std. dev. 0.201087
D 9.533889 1 9.533889 235.7774 <0.0001 PRESS 1.490334
B2 40.02204 1 40.02204 989.7631 <0.0001 Adj R2 0.994211
D2 6.835372 1 6.835372 169.0419 <0.0001 Adeq precision 92.71721
AB 3.900625 1 3.900625 96.46422 <0.0001
Residual 0.889591 22 0.040436
Lack of fit 0.823258 17 0.048427 3.650273 0.0789
Pure error 0.066333 5 0.013267
Cor total 202.5608 29
Model 4491.044 7 641.5777 20.89347 <0.0001
A 220.5 1 220.5 7.18075 0.0137 Mean 35.75
B 171.125 1 171.125 5.572816 0.0275 R2 0.869246
C 153.125 1 153.125 4.986632 0.0360 Std. dev. 5.541399
D 351.125 1 351.125 11.43465 0.0027 PRESS 1225.875
A2 668.9032 1 668.9032 21.78334 0.0001 Adj R2 0.827642
D2 259.8407 1 259.8407 8.461911 0.0081 Adeq precision 13.42728
BD 395.0156 1 395.0156 12.86398 0.0016
Residual 675.5561 22 30.7071
Lack of fit 583.4561 17 34.32095 1.863244 0.2542
Pure error 92.1 5 18.42
Cor total 5166.6 29
predict that if the residuals follow a normal distribution, as shown in Fig. 3, then the points will fall along a straight line for each case. However, some scattering is expected even with normal data; thus, the data can be considered to be normally distributed in the responses of certain models.
4.3. Three-dimensional plots of the regression and Optimization process:
To assess the interactive relationships among the independent variables and the responses of certain models, 3D surface response plots are created by Design Expert 6.0.7 (Fig. 4). The maximum TOC and COD removal efficiencies are 15.48%, 48.42%, respectively. The TiO2 and pH are the two main factors that improve the TOC and COD removal. The actual factors for removing these parameters include C0 of COD and RT. The current work reveals that photo-
catalyst of TiO2 in an AOP is more efficient in direct reaction with the petroleum wastewater treatment, achieving the highest removal rates for TOC and COD at alkaline conditions (pH = 8).
Optimization is performed to determine the optimum values of TOC and COD removal efficiencies using Design Expert 6.0.7. According to the optimization step, the desired goal for the operational conditions (C0 of COD, reaction time and pH) is chosen as "within" the range, while (TiO2 dosage) is chosen as "within" the minimum range to reduce the treatment cost. The responses (TOC and COD removal) are defined as "maximum" to achieve the highest performance. The program combines individual desirability into a single number and then searches to optimize this function basing on the response goal. Approximately 15.5% TOC and 48.5% COD removal rates are predicted by the models under optimized operational conditions (Table 6).
Table 5
Final equation in terms of coded and actual factors.
Final equation in TOC removal (%) = 14.8 + 0.61A + 0.87B + 1.1C + 0.7D - 3.4B2 - 1.4D2 - 0.5AB
terms of coded COD removal (%) = 49.3 - 3.5A + 3.1B + 2.9C + 4.4D - 13.9A2 - 8.7D2 + 4.9BD
Final equation in TOC removal (%) = -91.5 + 6.15TiO2 + 24 pH + 0.003TOD + 0.11RT - 1.5(pH)2 - 0.0004(RT)2 - 0.6TiO2pH
terms of actual COD removal (%) = -18.3 + 104.5TiO2 - 4.6 pH + 0.007COD + 0.2RT - 55.7(TiO2)2 - 0.002(RT)2 + 0.05(pH)RT factors
Normal Plot of Residuals
X Studpntized Residuals Y: Normal % Probability
(a) TOC removal
Normal Plot of Residuals
n i i r~
-1.42 -0.54 0.34 1.22 2.10
X Sturiflnti7Ad Residuals Y: Normal % Probability
(b) COD removal
Fig. 3. Design expert plot, normal probability plot of the studentized residuals for (a) TOC removal and (b) COD removal.
Fig. 4. Response surface models for TOC & COD removal efficiencies for a given TiO2 dosage (0.6 g/L), C0 of COD (1600 ppm), pH (8), reaction time (139 min).
5. Conclusion
In general, The value of COD consider more accurate than the value of TOC in checking of efficiencies of treatment in wastewater because the value of COD measure the oxygen required to oxidize the organic compounds in wastewater while the value of TOC measure amount of the carbon in wastewater (including dissolved carbon dioxide and carbon compounds such as amine and aromatic compounds so on). So, this technique seems efficient in reducing the organic contaminants because the achieved removal efficiency of COD is approximately 574 ppm as shown in Fig. 5. In addition, some reasons might be the relatively high COD and TOC values, the nature of organic compounds, and the photo-catalytic process.
The desirability function is found to be 0.903 for these optimum conditions. An additional experiment is then performed to confirm these optimum results, revealing agreement with the predicted response values as shown in Table 6.
In the present study, the performance of employing of solar photo-catalyst of TiO2 in the AOP on degradation of TOC and COD from petroleum wastewater at Sohar Oil Refinery in Oman is investigated. A central composite design (CCD) with response surface methodology (RSM) is applied to evaluate the relationships among operating variables, such as TiO2 dosage, C0 of COD, pH, and reaction time, to identify the optimum operating conditions. Quadratic models for the following two responses prove to be significant with
Table 6
Maximum TOC and COD removal efficiencies for model response and verification experiments under optimum conditions [TiO2 dosage (0.6 g/L), C0 of COD (1600ppm), pH (8) and reaction time (139 min)].
Unit Selected solution Lab experiments
TOC removal % 15.48 16.5
COD removal % 48.42 45
Desirability - 0.903098 -
■ The concentration before treatment (ppm)
■ The concentration after treatment (ppm)
190 156
1400 1200 1000 800 600 400 200 0
Fig. 5. The maximum TOC and COD removal efficiencies under the optimum conditions.
very low probabilities (<0.0001): chemical oxygen demand (COD) and Total Organic Carbon (TOC). The obtained optimum conditions include a TiO2 dosage (0.6 g/L), C0 of COD (1600 ppm), pH (8), reaction time (139 min). TOC and COD removal rates are 15.5% and 48.5%, respectively.
The predictions correspond well with experimental results (TOC and COD removal rates of 16.5%, and 45%, respectively). The solar photo-catalyst of TiO2 of high COD wastewater is a unique treatment process utilizing renewable solar energy and treating the wastewater with minimum chemical input.
References
[1] Augustine Chioma Affam, Malay Chaudhuri, Degradation of pesticides chlorpyrifos, cypermethrin and chlorothalonil in aqueous solution by TiO2 photocatalysis, J. Environ. Manag. 130 (2013) 160—165.
[2] L.N. Mota, L.F. Albuquerque, L.T.C. Beltrame, O. Chiavone-Filho, A. Machulek Jr., et al., Advanced oxidation processes and
their application in the petroleum industry: a review, Braz. J. Petroleum Gas 2 (2008) 122—142.
[3] Chandan Singh, Rubina Chaudhary, Kavita Gandhi, Preliminary study on optimization of pH, oxidant and catalyst dose for high COD content: solar parabolic trough collector, Iran. J. Environ. Health Sci. Eng. (2013) 10—13.
[4] Jose Fenoll, Encarnacn Ruiz, Pilar Helln, Pilar Flores, Sin Navarro, Heterogeneous photocatalytic oxidation of cypro-dinil and fludioxonil in leaching water under solar irradiation, Chemosphere 85 (2011) 1262—1268.
[5] Ramanjot Kaur Sidhu, Amit Dhir, Studies on the Solar Assisted Advanced Oxidative Treatment of Procion Dye, A Dissertation report for the award of Master of Technology Degree from Thapar University, 2011.
[6] F.V. Santos, E.B. Azevedo, G.L. Sant'Anna Jr., M. Dezotti, Photo-catalysis as a tertiary treatment for petroleum refinery waste waters, Braz. J. Chem. Eng. 23 (04) (2006) 451—460.
[7] S. Shanmuga Priya, M. Premalatha, N. Anantharaman, Solar photo-catalytic treatment of phenolic wastewater potential challenges and opportunities, ARPN J. Eng. Appl. Sci. 3 (6) (2008). ISSN 1819—6608.
[8] Javad Saien, Fatemeh Shahrezaei, Organic pollutants removal from petroleum refinery wastewater with nanotitania photo-catalyst and UV light emission, Int. J. Photoenergy (2012) 5. ID 703074.
[9] M.Y. Noordin, V.C. Venkatesh, S. Sharif, S. Elting, A. Abdullah, Application of response surface methodology in describing the performance of coated carbide tools when turning AISI 1045 steel, J. Mater. Process Technol. 145 (1) (2004) 46—58. http://dx. doi.org/10.1016/S0924-0136(03)00861—6.
[10] D.C. Montgomery, Design and Analysis of Experiments, seventh ed., 2008.
[11] S. Salem, Abu Amr, Hamidi Abul Aziz, Mohd Nordin Adlan, Mohammed J.K. Bashir, Optimization of semi-aerobic stabilized leachate treatment using ozone/Fenton's reagent in the advanced oxidation process, J. Environ. Sci. Health A 48 (2013) 720—729.
[12] A.M. Joglekar, A.T. May, Product excellence through design of experiments, Cer. Foods World 32 (1987) 857—868.
