Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010

Fenton Oxidation of Natural Gas Plant WastewaterAbdul Aziz Omar, Raihan Mahirah Ramli, and Putri N. Faizura Megat Khamaruddin

AbstractWastewater produced in a natural gas plant treatment contains organic compounds namely as alkanolamines. These organic compounds are not biodegradable which makes it resistant to the current wastewater treatment facility. Advanced oxidation process which involves reactive hydroxyl radical is one of the options to treat this type of wastewater. As much as 73% of COD reduction was achieved with initial pH of 3 and H2O2/Fe2+ molar ratio of 10. The proposed rate constant for this wastewater taken into account only the single highest concentration substrate and the applicability of the rate constant to predict final COD value has been proved. Index Termsdegradation, diisopropanolamine, oxidation, chemical oxygen demand Fentons

aw natural gas extracted from reservoir contains impurities that must be removed before it can be used. The raw natural gas which primarily consists of methane also contains varying amounts of other components such as heavier gaseous hydrocarbons (ethane, propane, butane etc.), acid gases (carbon dioxide, hydrogen sulfide, mercaptans etc.), water (vapor and liquid) and liquid hydrocarbons. After the treatment, the processed natural gas will contain almost pure methane and it is used as fuel by residential, commercial and industrial consumers. Amine gas treating, also known as gas sweetening and acid gas removal, refers to a group of processes that use aqueous solutions of various alkanolamines to remove hydrogen sulfide (H2S) and carbon dioxide (CO2) from natural gas. This treating unit is commonly used in refineries, petrochemical plants, natural gas processing plants and also some other industries. Development of alkanolamines as absorbent for acidic gases was discovered by R. R. Bottoms in 1930. The first alkanolamine commercially available for acid gas treatment was triethanolamine (TEA). Since then, the other members of alkanolamines were introduced in the market and evaluated as possible acid gas absorbents [1]. Various amines are now available such as monoethanolamine (MEA), diethanolamine (DEA), diisopropanolamine (DIPA), methyldiethanolamine (MDEA) and TEA.Manuscript submitted 15 December 2009. Abdul Aziz Omar is with Chemical Engineering Department, Universiti Teknologi PETRONAS, Tronoh, Perak, Malaysia (phone: 605-3687090; fax: 605-3655670; e-mail: aaziz_omar@ petronas.com.my). Raihan Mahirah Ramli is now with Chemical Engineering Department, Universiti Teknologi PETRONAS, Tronoh, Perak, Malaysia (e-mail: raihan_mahirah@yahoo.com). Putri Nadzrul Faizura Megat Khamaruddin is with Chemical Engineering Department, Universiti Teknologi PETRONAS, Tronoh, Perak, Malaysia (e-mail: nadzru@petronas.com.my).

R

I. INTRODUCTION

Despite wide application of alkanolamines, treatment technology to degrade these organic compounds is still need to be developed. In the past few decades, advanced oxidation processes (AOPs) have attracted many researchers due to its ability to degrade wide range of recalcitrant organics that is impossible to be degraded by other methods. Compared to other wastewater treatment processes such as biological, classical chemical and physical-mechanical, AOPs have proved their abilities to mineralize organic contaminants present in wastewater. They have been used to degrade 99.25% of azo dye Amido black [2], 90% of cork cooking [3], and 87.3% of pharmaceutical waste [4]. However, the research on applicability of AOPs to degrade alkanolamines has yet to be conducted. Very little research on amine itself is available and none is on alkanolamines. Glaze et al [5] defined AOPs as near ambient temperature and pressure water treatment processes which involve the generation of hydroxyl radicals in sufficient quantity to effect water purification. The main feature of AOPs is hydroxyl radical (OH) which has a high oxidation potential and acts rapidly with most organic compounds. Among the AOPs, Fentons oxidation is the simplest process which involves only ferrous catalyst and hydrogen peroxide. Fentons reagent had been applied to degrade a wide range of contaminants, mainly recalcitrant organic pollutants. It was first reported by Fenton [6] in 1876. Since then, the applicability of Fentons reagent to the degradation of many recalcitrant organic compounds has been reported and the research is still going. Kavitha and Palanivelu [7] reported that destruction of cresols can be done by using Fenton oxidation, where the degradation efficiency was as high as 82%. In textile industry, the reagents used are very diverse in chemical compositions which are not easily amenable by biological treatment. Among the compounds found by GC-MS are 3-methylbenzoic acid, ethyl ester 4-ethoxy-benzoic acid, and 2,8dimethyl-4-methylene-nonane [8]. Treatment of synthetic wastewater in laboratory containing azo dyes was studied by Sun et al. [2] by using Fentons reagent. Under optimal conditions (pH = 3.50, [Fe2+]o:[H2O2]o = 1:10, [dye]o = 50 mg/L, temperature = 25C), 99.25 percent degradation efficiency of dye was achieved. Agriculture industry produces large volumes of livestock wastewater which will be a problem to the conventional wastewater treatment system if the strength of the wastewater is high. Heavy contamination of livestock wastewater is characterized by high suspended solids, high COD, high BOD and high nitrogen, which affected the surface water if it is not treated [9]. A study by Lee and Shoda [10] using livestock wastewater from Japan revealed that the wastewater was characterized as high COD using chromate (CODcr of 5,000 to 5,700 mg/L), high total solid (5,126 to 5,900 mg/L), pH in the range of 8.4 to 8.7, and the color was dark grey. The treatment of the wastewater by Fentons reagent has reduced more than 80 percent of 1

Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010 CODcr and 95 percent of color with only single addition of reagents under optimal condition. II. EXPERIMENTAL A. Materials Hydrogen peroxide (for synthesis, 30% w/w solution) was purchased from Merck. Iron (II) Sulfate 7-hydrate (FeSO4.7H2O) used was from HmbG Chemicals, whereas concentrated sulphuric acid, H2SO4 (98%) was from Systerm and sodium hydroxide, NaOH (1N) was from R&M Chemicals. Chemical oxygen demand (COD) was determined in a HACH DR5000 spectrophotometer based on dichromate standard procedure. Table 1 shows the characteristics of the wastewater collected from a natural gas plant in Malaysia.Table 1: Characteristics of the wastewater

continuously using Masterflex peristaltic pumps. The dosing flowrate was calculated based on the dosage determined divided by reaction time. III. RESULT AND DISCUSSION A. Effect of wastewater pretreatment According to Table 1, the wastewater has high oil and suspended solid content which may interfere during Fentons oxidation. A set of experiments were conducted to investigate the effect of oil and suspended solid in the wastewater to the degradation efficiency. Three different wastewater samples were used to study the effect; pretreated sample, partial treated sample (contains suspended solid only) and untreated sample (contains oil and suspended solid). For initial study, theoretical amount of hydrogen peroxide was used (216 mL/L) which was calculated based on the initial COD of the wastewater sample. Molar ratio of H2O2/Fe2+ of 8 and initial pH of 2 was used as a continuation from the previous research work. The experiments were conducted in ambient temperature and pressure.70 60

Characterization Study pH Chemical Oxygen Demand 5-day Biological Oxygen Demand Total Organic Carbon Total Suspended Solid Total Volatile Solid Total Nitrogen Oil and Grease

Amount 8.67 17, 028 mg/L 909 mg/L

4,517 mg/L 640 mg/L 916 mg/L 622 mg/L

COD removal (%)

3,337 mg/L

50 40 30 20 10 0Pretreated sample Partial treated sample Untreated sample

B. Experimental Procedure Experiments were conducted in batch mode in a one-liter jacketed glass reactor with provisions for sampling, temperature and pH probes. The reactor was placed on a magnetic stirrer. Water was passed through the jacket during the reaction in order to maintain the solution temperature at 30C. In general the experiments were carried in 30 minutes. Volume of solution tested per batch was 500 ml. The solutions pH was corrected before mixing in calculated volume of FeSO4.7H2O solution and 30% H2O2. During the experimental run, 1 ml of reacting solution was withdrawn from the reactor at different time intervals. The sample was immediately treated with 4 ml of 1 N NaOH to stop the oxidation process. Residual H2O2 must be removed as it will interfere with the COD measurement. Increasing the pH and heating the solution higher than 40C will degrade H2O2. Thus samples were heated up to about 70C for 30 minutes [7, 11-12]. Upon 30 minutes and when no bubbles were observed in the samples, the samples were left to cool to room temperature. Samples were filtered using Whatman Puradisc Aqua 30 syringe filter, 0.45m to remove the ferric oxide precipitate for COD measurement. In the previous research work, continuous addition of reagents was found to be the best way to dose the reagents. Therefore, in this work, the reagents were dosed 2

Fig. 1. COD degradation effect with different wastewater pretreatment

Fig. 1 shows the effect of wastewater pretreatment before Fentons oxidation to the degradation efficiency. The pretreated wastewater gave higher COD reduction compared to the other wastewater. High amount of oil and suspended solid in the untreated and partial treated wastewater slightly reduce the effectiveness of Fentons oxidation that only oxidized soluble organic compounds inside the wastewater. Even though the percentage removal has only slight differences, the removal of these two pollutants is very important to adhere with the Environmental Quality Acts (1974) whereas stated in the acts the discharge limit for oil and suspended solid into inland waters is 10 and 100 mg/L respectively [13]. Besides, the presence of oil and suspended solid during Fentons system may cause plugging in the equipment and regular maintenance may be required to clean up the equipment. B. Effect of different molar ratio Wide range of H2O2/Fe2+ molar ratio was reported in the literature. The optimum molar ratio of reagents is very important to ensure the treatment is economically feasible. Besides, excess reagents will defer the degradation

Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010 efficiency. A study by Casero et al. [14] on the degradation of amine by Fentons reagent was used as the reference where they reported a range of optimum molar ratio 5 to 40. Since the target substrate is an amine, a set of experiments were conducted at different molar ratio of 5, 10, 20 and 30. The experiments were conducted at initial pH of 2, constant amount of hydrogen peroxide was used (216 mL/L), and at ambient temperature and pressure. The amount of FeSO4.7H2O was varied following the required molar ratio.70 60

Thus, a series of experiment were conducted at different pH values of 2, 3, 4, and 5. The H2O2/Fe2+ molar ratio of 10 was used. The process was carried out at ambient temperature and pressure.

80 70 60

COD removal (%)

50 40 30 20 10 0

COD removal (%)

50 40 30 20 10 0 5 10 20 30

pH 2

pH 3

pH 4

pH 5

Fig. 3. COD degradation effect with different solutions initial pH

H2O2/Fe2+ molarFig. 2. COD degradation effect with different molar ratio

In Fig 2, the highest COD percentage removal was obtained at molar ratio of 10 with 64% COD removal. When the ratio was reduced, the COD removal dropped to 50 percent. The similar result was achieved at molar ratio of 20 and the COD removal dropped further with higher molar ratio of 30. At lower ratio, the amount of Fe2+ in the system was not sufficient to generate required amount of hydroxyl radicals for substrate oxidation (reaction 1). In contrast, at higher molar ratio (20 30), excess Fe2+ led to the scavenging reaction (2). The hydroxyl radicals react with ferrous ion instead of attacking the organic substrate. The optimum H2O2/Fe2+ molar ratio for the treatment of the wastewater was 10. Thus, this ratio was used in the following study. Fe2+ + H2O2 Fe3+ + OH

During the reaction, the initial pH of the sample decreased from the pH that was set initially. Formation of carboxylic acids such as acetic acid, oxalic acid and formic acid ([18]) may be the reason behind this observation. From Fig 3, the residual COD of the wastewater was at the lowest value at pH of 3 (73% COD removal). Meanwhile, at the highest pH value, only 20 percent of COD removal was achieved. This result showed that Fentons oxidation is highly affected by the initial pH of the wastewater where it is only effective in acidic environment as suggested by reaction (3). From this study, optimum degradation was achieved at pH 3 and it started to reduce at both lower and higher pH value (Fig. 3). The explanation behind this phenomena was that at low pH value (4) was used, the treatment efficiency decreased due to the formation of ferric-hydroxo complexes ([18], [20-22]) which deactivated the ferrous catalyst supplied to the system leading to the reduction of OH radicals generation. Besides, at this pH value, hydrogen peroxide as the oxidant decomposed into oxygen and water instead of hydroxyl radicals. D. Effect of hydrogen peroxide concentration Besides ferrous sulfate concentration, hydrogen peroxide amount is another main factor that directly affects the degradation of organic compounds by Fentons oxidation. As the source of hydroxyl radicals (the essential component in the reactions), the correct concentration of hydrogen peroxide is vital either in treatment efficiency or from economic view. Effect of hydrogen peroxide to the degradation efficiency was studied by varying that amount of it used in the experiment. This moment, the concentration of hydrogen peroxide was either increased or decreased 20 3

Previous study on the application of Fentons oxidation to degrade various organic pollutants reported that the optimal pH range was between 2.5 to 4.5 ([9], [16-18]).

Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010 percent for each experiment (86 to 260 mL/L). Even though optimum molar ratio of 10 was already achieved in previous experiments, this study is just to reconfirm as well as to study the effect of hydrogen peroxide to the oxidation process. Ferrous sulfate used was 47.95 g/L with initial pH of 3 at ambient conditions.80 70 60

OH + Fe2 + k 9 Fe3+ + OH

(9)

Quite a few species involved in the reaction scheme are not present at the beginning and since OH is the primary oxidizing species in the overall process, we consider the reactions (6), (7), (8) and (9) only in this simplified analysis. A pseudo-steady state balance of the rates of generation and disappearance of the OH radicals leads to following expression for its concentration.

COD removal (%)

50 40 30 20 10 0 86 129 17 216 26

] k [ OH] = k [H O [H+ O [S[Fe k ][Fe ] ] k ]+ 2+ 6 2 2 2+ 7 2 2 8 9

(10)

where k6, k7, k8 and k9 are the rate constants for reaction (6), (7), (8) and (9) respectively. The initial rate of mineralization of the substrate can be written asVolume of H2O2 (mL/L)

[rs ]0 = k8 [S]0 [ OH ]0 = k8 [S]0

k 6 [H 2 O 2 ]0 Fe 2 + 0 k 7 [H 2 O 2 ]0 + k8 [S]0 + k 9 Fe 2 +

[

]

[

]

0

Fig. 4. COD degradation effect with different volume of hydrogen peroxide

(11) where the subscript 0 denotes zero time. The equation can be rearranged to k 6 [H 2 O 2 ]0 Fe 2 + [rs ]0

Comparison of the degradation percentage in Fig. 4 showed that the highest COD removal was obtained at the theoretical concentration of hydrogen peroxide (216 mL/L). These findings confirmed the results that we have obtained in the previous set of experiments. At higher concentration, the removal of COD decreased. The same findings were observed at the concentration lower than that where COD removal decreased when concentration of hydrogen peroxide was reduced. At higher concentration, this phenomenon happened due to the scavenging effect of hydrogen peroxide (reaction 5).

[

] [S] [S] = 1 {k [H O ] + k [Fe ] } k0 0 2+ 0

7

2

2 0

9

0

8

(12)

Y=

1 X k8

(13)

OH + H2O2 H2O + HO2

(5)

On the other hand, insufficient hydroxyl radicals caused the degradation efficiency decreased at lower concentration of hydrogen peroxide due to insufficient amount of hydroxyl radicals (reaction 1). Due to that, the theoretical concentration of hydrogen peroxide (216 mL/L) is required in order to get the highest COD removal. This finding hereby confirmed that the optimal H2O2/Fe2+ molar ratio for the treatment of this wastewater is 10.

IV. A SIMPLIFIED RATE MODEL FOR MINERALIZATION In order to develop a rate equation, the following steps of generation and reaction of OH groups are proposed. Here S stands for the substrate.

H 2O 2 + Fe2 + k 6 OH + Fe(OH)2 +

(6) (7) (8)

H 2O 2 + OH k 7 HO 2 + H 2O S+ OH k 8 degradation products

The above equation can be used to determine the degradation rate constant k8 using the experimental data on the rate of degradation when only the concentration of added H2O2 is varied keeping constant those of substrate (S) and of Fe2+. The rate constant k6, k7 and k9 for the reaction (6), (7) and (9) respectively are available in the literature ([17-18], [23]). We have taken the following values of the above rate constants: k6 = 70, k7 = 3107 and k8 = 3108 M1 -1 s respectively. A plot of the quantity Y against X [see Eq. (13)] should produce a straight line passing through the origin with a slope equal to the inverse of the constant, k8. The plot of the experimental data in the form of equation (13) gives straight lines as shown in Fig. 5. From the slope of the line, the rate constant for mineralization is estimated to be k8 = 1.43107 M-1min-1 = 2.38105 M-1s-1. It is to be noted that we have taken the calculated rate of degradation as the rate of removal of COD or, in other words, the rate of complete oxidation of the substrate. Since the wastewater sample contains traces of sulfinol (a solvent mixture of 50 % DIPA, 25 % sulfolane and 25 % water) and other unrecognized compounds with smaller concentration, the substrate concentration was calculated based on the concentration of DIPA. It may be considered to be a lumped representation of the process of degradation of the substrate as well as the intermediates. Although the degradation data

4

Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010 fitted reasonable well in the above model, the rate constant for the wastewater was only based on the concentration of DIPA alone. Development of a model suitable for the wastewater by taking into account other compounds may perhaps be taken up as an extension of this work. The experimental data on COD removal for same initial COD concentration (17000 mg/L) and ferrous sulfate dose (23.97g) in 500 ml reaction mixture but with different H2O2 dose [43, 64.5, 86, 108 and 130 ml; 30% w/w) as shown in Fig. 4 are used to calculate the initial rate of degradation. The reduction of COD over thirty minutes was used for calculation of [rs]0.20

V. CONCLUSION The application of Fentons oxidation to degrade natural gas plant wastewaters which primarily consist of nonbiodegradable organic compounds has been studied and proved to be efficient. With initial COD concentration of approximately 17,000 mg/L, the optimum H2O2/Fe2+ molar ratio is found to be at 10 and wastewaters initial pH of 3. The highest COD removal achieved was 73% which can be further reduced in biological treatment system due to the formation of carboxylic acids. The proposed rate constant fitted reasonably well and found to be applicable to predict the final COD at different conditions. ACKNOWLEDGEMENT

16

The authors would like to express their gratitude to Universiti Teknologi PETRONAS for funding the research. REFERENCES[1] Kohl, A. I., and Nielsen, R. B., Gas purification, Fifth Edition, Gulf Publishing Company, Houston, Texas, 1997 [2] Sun, J. H., Sun, S. P., Wang, G. L., and Qiao, L. P., Degradation of azo dye Amido Black 10B in aqueous solution by Fenton oxidation process, Dyes and Pigment B136, 2006; 258-265G [3] Guedes, A. M. F. M., Madeira, L. M. P., Boaventura, R. A. R., and Costa, C. A. V., Fenton oxidation of cork cooking wastewater-Overall kinetic analysis, Water Res. 37, 2003; 30613069 [4] Tekin, H., Bilkay, O., Ataberk, Selale S., Balta, Tolga H., Ceribasi, I. Haluk, Dilek, Sanin F., Dilek, Filiz B., and Ulku, Yetis, Use of Fenton oxidation to improve the biodegradability of a pharmaceutical wastewater, Journal of Hazardous Material B136, 2006; 258-265 [5] Glaze, W. H., Kang, J. W., and Chapin, D. H., The chemistry of water treatment processes involving ozone, hydrogen peroxide and ultraviolet radiation, Ozone Science Engineering 3, 1987; 145-152 [6] Matthew, A. T., Fenton and modified Fenton methods for pollutant degradation; In Chemical degradation methods for wastes and pollutants, Marcel Dekker Inc., New York, 2003; pp. 164-194 [7] Kavitha, V., and Palanivelu, K., The role of ferrous ion in Fenton and photo-Fenton processes for the degradation of phenol, Chemosphere 55, 2005; 1235-1243 [8] Perez, M., Torrades, F., Domenech, X., and Peral, J., Fenton and photo-Fenton oxidation of textile effluents, Water Res. 36, 2002; 2703-2710 [9] Shin, J. H., Lee, S. M., Jung, J. Y., Chung, Y. C., and Noh, S. H., Enhanced COD and nitrogen removals for the treatment of swine wastewater by combining submerged membrane bioreactor (MBR) and anaerobic upflow bed filter (AUBF) reactor, Process Biochemistry 40, 2005; 3769-3776. [10] Lee, H., and Shoda, M., Removal of COD and Color from livestock wastewater by Fenton method, Journal of Hazardous Materials, 2007; doi: 10.1016/j.jhazmat.2007.09.097 [11] Lou, J. C., and Lee, S. S., Chemical oxidation of BTX using Fentons reagent, Journal of Hazardous Material 12, 1995; 185193 [12] Jones, C. W., Introduction to the preparation and properties of hydrogen peroxide; In: Application of hydrogen peroxide and derivatives, Royal Society of Chemistry, Cambridge, UK, 1999; 30 [13] Environmental Quality Act (EQA) 1974 [Act 127], Environmental Quality (Sewage and Industrial Effluents) Regulations 1979 [14] Casero, I., Sicilia, D., Rubio, S., and Perez-Bendito, D., Chemical degradation of aromatic amines by Fentons reagent, Water Res. 31, 1997; 1985-1995

12

Y8

4

0 0 2e+ 4e+ 6e+ 8e+ 1e+

XFig. 5. Plot of Y vs. X for equation (13)

Although the degradation data for the wastewater fitted reasonable well in the above model, verification is necessary to ensure the rate constant is applicable at different condition of oxidation. The applicability of this kinetic expression was verified by using the rate constant obtained from Fig. 5 to calculate the predicted final COD. Several experiments at different conditions (different initial COD and reagents amount) were carried out and the results obtained were summarized in Fig. 6.Calculated data

800

Experimental data

COD residue (mg/L)

600

400

200

0

Run

Run

Run

Run

Run

Fig. 6. Applicability of the rate constant to the wastewater at different experimental conditions

It was observed in Fig. 6 that there are only small differences between the calculated final COD with the final COD from the experiments with an average between 5 percent. Therefore, we concluded that the rate constant obtained from this study is valid to be used for this particular wastewater sample.

5

Canadian Journal on Chemical Engineering & Technology Vol. 1, No. 1, February 2010[15] Walling, C., Fentons reagent revisited, Acc. Chem. Res. 8, 1975; 125 [16] Azbar, N., Yonar, T., and Kestioglu, K., Comparison of various advanced processes and chemical treatment methods for COD and color removal from a polyester and acetate fiber dyeing effluent, Chemosphere 55, 2004; 35-43 [17] Neyens, E., and Baeyens, J., A review of classic Fentons peroxidation as an advanced oxidation technique, Journal of Hazardous Material 98, 2003; 33-50 [18] Burbano, A. A., Dionysiou, D. D., Suidan, M. T., and Richardson, T. L., Oxidation kinetics and effect of pH on the degradation of MTBE with Fenton reagent, Water Res. 39, 2005; 107-118 [19] Kwon, B. G., Lee, D. S., Kang, N., and Yoon, J., Characteristics of -chlorophenol oxidation by Fentons reagent, Water Res. 33, 1999; 2110-2118 [20] Kuo, W. G., Decoloring dye wastewater with Fentons reagent, Water Res. 26, 1992; 881-886 [21] Y.W. Kang, K.Y. Hwang, Effects of reaction conditions on the oxidation efficiency in the Fenton process, Water Res. 34, 2000; 27862790. [22] S.M. Kim, S.U. Geissen, A. Vogelpohl, Landfill leachate treatment by a photoassisted Fenton reaction, Water Sci. Technol. 35, 1997; 239248. [23] Kang, S. F., Wang, T. H., and Lin, Y. H., Decolourization and degradation of 2,4-dinitriphenol by Fentons reagent, J. Environ. Sci. Health A34, 1999; 301-307

6