Water Research 37 (2003) 1170–1176

Pretreatment of Afyon alcaloide factory’s wastewater by wet air oxidation (WAO) Y. Kac-ara, E. Alpayb,*, V.K. Ceylanc a

Engineering Faculty, Department of Environmental Engineering, Akdeniz University, 07200, Topc-ular, Antalya, Turkey b ! Engineering Faculty, Department of Chemical Engineering, Ege University, 35100 Bornova, Izmir, Turkey c Us-ak Engineering Faculty, Department of Chemical Engineering, Afyon Kocatepe University, 64300 Us-ak, Turkey Received 7 May 2002; received in revised form 8 August 2002; accepted 7 September 2002

Abstract In this study, pretreatment of Afyon (Turkey) alcaloide factory wastewater, a typical high strength industrial wastewater (chemical oxygen demand ðCODÞ ¼ 26:65 kg m3 ; biological oxygen demand ðBOD5 Þ ¼ 3:95 kg m3 Þ; was carried out by wet air oxidation process. The process was performed in a 0:75 litre specially designed bubble reactor. Experiments were conducted to see the advantages of one-stage and two-stage oxidation and the effects of pressure, pH, temperature, catalyst type, catalyst loading and air or oxygen as gas source on the oxidation of the wastewater. In addition, BOD5 =COD ratios of the effluents, which are generally regarded as an important index of biodegradability of a high-strength industrial wastewater, were determined at the end of some runs. After a 2 h oxidation (T ¼ 1501C; P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 7:0), the BOD5 =COD ratio was increased from 0.15 to above 0.5 by using the salts of metals such as Co2þ ; Fe2þ ; Fe2þ þ Ni2þ ; Cu2þ þ Mn2þ as catalyst. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: COD removal; Homogeneous catalysis; Nonbiodegradable wastes; Wastewater treatment; Wet air oxidation (WAO)

1. Introduction Organic compounds are used in the manufacture of a wide variety of commercial chemical products. The use of these organic compounds in a manufacturing process invariably results in different types of wastewater that contains significant amounts of organic compounds. Discharge of these wastewaters without treatment into a natural water body is unacceptable in practice since the discharge upsets water quality of the receiving water body. In addition to the potential toxicity of the organic compounds, the dissolved oxygen (DO) concentration of the receiving water body polluted by untreated wastewater can fall below the level necessary for maintaining normal aquatic life. Hence, increasingly stringent *Corresponding author. Tel.: +90-232-3880-016; fax: +90232-3741-401. E-mail address: [email protected] (E. Alpay).

restrictions are being imposed by governments on the concentration of these organic compounds in the wastewater for safe discharge. The choice of the method for the treatment of such wastes is governed by various factors such as the constituents (organic or inorganic), their concentration, volume to be treated, and toxicity to microbes. The various treatment methods available are chemical treatment (ozonation, wet oxidation, UV irradiation, etc.), physical treatment (adsorption, reverse osmosis, etc.), biological treatment, incineration, etc. More often combination of the above methods is used to get better results [1]. Among the various types of processes, which can be used for treating aqueous wastes polluted with organic matter, wet air oxidation (WAO) is very attractive. It is an enclosed process with very limited interaction with the environment. It can be coupled with a biological treatment facility to eliminate the sludge, or treat any kind of waste, even toxic. The basic idea of WAO is to

0043-1354/03/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 4 4 8 - 7

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

enhance contact between molecular oxygen and the organic matter to be oxidized [2]. It accomplishes oxidation at elevated temperatures (125–3201C) and pressures (0.5–20 MPa) using a gaseous source of oxygen (usually air). Temperature is recognized as the most important process parameter. The pressure is also required to keep water in the liquid state. Water which makes up the bulk of the aqueous phase serves to modify the oxidation reactions so that they proceed at relatively low temperatures in the range of 125–3201C; and at the same time serves to moderate the oxidation rates removing excess heat by evaporation. Water also provides excellent heat transfer medium that enables efficient heat transfer. The oxidation products may be inorganic salts, simpler forms of biodegradable compounds or may be carbon dioxide and water [3]. In most applications, WAO is not used as a complete treatment method, but only as a pretreatment step where the wastewater is rendered to nontoxic materials and the COD is reduced for the final treatment. For integrated WAO–biological treatment process, more detailed studies concerning the WAO pretreatment step are necessary for the design of a rational and efficient integrated process [4]. The WAO process has been subjected to numerous investigations by researchers in the past decades as a pretreatment step before the biological treatment [5–8]. The purpose of this study is to investigate the pretreatment of the wastewater of alcaloide factory in Afyon (Turkey) by WAO experimentally and to find the effects of various operating parameters on the efficiency in order to optimize the WAO treatment process. Also considered in the present work is the determination of the change of the ratio of biological oxygen demand to chemical oxygen demand. Such a ratio is generally regarded as an important index of biodegradability of the treated high-strength industrial wastewater.

1171

The colour of wastewater was light brown. Before it was loaded into the reactor, it was filtered by a filter paper. 2.2. Catalysts It is well known that metal salts act as effective catalyst in WAO processes [9–11]. Thus, experiments were performed using four different catalysts, namely, NiðNO3 Þ2 ; FeCl2 ; CuðNO3 Þ2 and CoðNO3 Þ2 and two mixed catalysts: CuðNO3 Þ2 þ MnCl2 and FeCl2 þ NiðNO3 Þ2 ; in the form of metal salt solution in order to promote the oxidation of organic compounds present in the wastewater. The reason for testing mixed catalysts was to see the effect of synergy, if any. The amount of catalyst added to the reactor was 0:25 kg m3 in terms of metal ion concentration. The metal salts used as catalysts in all experiments were the products of Merck or Riedel de Ha!en. 2.3. Experimental setup The flow diagram of the experimental setup is shown in Fig. 1. The experimental set up consisted mainly of a reactor and a condenser. It was equipped with suitable PG

V3 FRI

V5 CW o

E V4

CW I V1

G

F P

TI

R V7

2. Material and methods

OO

2.1. Characteristics of wastewater used in the experiments

V6

OI

GS

The wastewater used in the experiments was obtained from alcaloide factory in Afyon, Turkey. This factory produces morphine from the capsule of opium for pharmaceutical industry. The amount of wastewater produced at this factory is about 180; 000 m3 yr1 : The wastewater contains mainly bioresistance organic compounds such as morphine, aniline, and toluol, etc. This wastewater had a COD value of 26:65 kg m3 ; BOD5 of 3:95 kg m3 and pH of 3.7. These characteristics of wastewater were observed to remain constant during the experimental program, which lasted for about 8 months.

V2

S

Fig. 1. Schematic diagram of the experimental setup. R ¼ reactor; E ¼ reflux condenser, G ¼ gas (air or oxygen) line, F ¼ charge (wastewater) line, S ¼ discharge line, GS ¼ gas sparger, FRI ¼ flow rate indicator, TI ¼ temperature indicator, PG ¼ pressure gauge, CWI ¼ inlet of cooling water, CWO ¼ outlet of cooling water, OI ¼ inlet of hot oil, OO ¼ outlet of hot oil, V1 ; V3 ¼ needle valves, V4 ; V5 ; V6 ; V7 ¼ gate valves, V2 ¼ ball valve, P ¼ plug:

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

1172

ðV7 Þ valves were opened. During this time, valves V1 and V3 were kept closed. When the reactor content reached the desired temperature, valves V1 and V3 were carefully opened to let air or oxygen slowly bubble into the reactor through the sparger and leave the reactor from the top continuously. After desired pressure and temperature were attained, this time was taken as the ‘‘zero’’ time for the reaction. During the reaction, temperature, pressure, gas flow rate were measured and recorded at 5 min intervals. At the end of the experiment, first valve V1 was closed and then the hot oil system was shut down. It was maintained about 2–2:5 h for the reactor content to cool down. Then, the reaction mixture was discharged into a glass flask by opening valve V2 : It was filtered by a filter paper. All the samples taken were analysed for COD and BOD5 by standard methods [12].

measuring devices, such as thermocouple, rotameter and pressure gauge. Based on the literature and experience surveys, the material of construction for reactor was chosen as titanium. Thus, the oxidation of the wastewater in these experiments was carried out in a 0:75 l titanium bubble reactor. The top of the reactor is connected to a reflux condenser with a stainless steel flange. The reactor was equipped with a heating jacket and a gas sparger. The gas (air or oxygen) entered the reactor through the sparger, which was a titanium tube of 10 mm outside diameter and in L form. There were nine holes of 1 mm diameter on the horizontal part of the sparger through which air or oxygen bubbled out at high speed and thus ensured proper agitation. The temperature of the reactor was measured with 70:11C accuracy by a 3.5 digit temperature indicator using thermocouple (iron–constantan). The thermocouple was long enough to measure the temperature of the liquid phase in the reactor. Heating was done by hot oil circulating in the jacket around the reactor. The required temperature of the reactor was maintained by adjusting the temperature and the flow rate of hot oil. In order to keep oxidation reaction in the liquid phase, the system pressure was controlled manually by a needle valve ðV3 Þ located in the exit gas line. Gas flow rate in the exit line was measured by a calibrated rotameter (Sho-Rate, Model 1355) and it was adjusted with a needle valve ðV1 Þ: The wastewater was filled to the reactor through charge line, which has a plug on top of it. On the discharge line, there was a ball valve ðV2 Þ:

3. Experimental results and discussion Firstly, the treatment of wastewater was performed using both one-stage and two-stage Loprox processes [13] in order to see which one is more advantageous for the oxidation of wastewater. In two-stage Loprox process, the first stage operation was conducted in an alkaline medium at a temperature of 1201C and at a pressure of 0:5 MPa with an airflow rate of 1:57  105 m3 s1 : The reaction time was selected as 2 h: In the subsequent second stage, the wastewater drawn from the first stage was used. The pH of the wastewater was adjusted to 2.0 by addition of sulphuric acid. To this solution, 0:50 kg m3 FeCl3 was then added as catalyst and oxidation reaction was conducted at a temperature of 1451C under a pressure of 0:5 MPa with an airflow rate of 1:57  105 m3 s1 : In one-stage process, same experiments were performed in only acidic solutions. The results given in Table 1, show that one-stage process is superior to two-stage process. In reality, while the COD removal is only 13.6% with catalyst in two-stage process, this is 31.7% under the same conditions in onestage process. After these results, it was decided to conduct the subsequent experiments using one-stage process.

2.4. Experimental procedure The operating procedure of wastewater treatment was as follows. Before each experiment, the reactor was cleaned with distilled water. This cleaning step was repeated three or four times until the discharged water was visibly clean. The sealing of the reactor was checked by introducing air at a pressure of 0:65 MPa to the reactor. 0:3 litre (about 40% the capacity of the reactor) of wastewater, together with the catalyst, if required, was charged into the reactor, and plug was closed tightly. It was maintained until the hot oil temperature reached the set value. Then hot oil inlet ðV6 Þ and outlet Table 1 The results obtained from two-stage and one-stage operations pH

t (h)

T ð1CÞ

PT (MPa)

FeCl3 ðkg m3 Þ

COD removal (%) 4.9 8.7 (13.6) 17.8 31.7

Two-stage process

1st stage 2nd stage

10.0 2.0

2.0 2.0

120 145

0.50 0.50

Not used 0.50

One-stage process

Without catalyst With catalyst

2.0 2.0

2.0 2.0

145 145

0.56 0.58

Not used 0.50

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

35 COD removal, %

30 25 20 15 10 P=0.65 MPa P=0.80 MPa

5 0 0

1800

3600 5400 Reaction time (s)

7200

9000

Fig. 2. Effect of pressure on the COD removal. Conditions: T ¼ 1501C; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 3:7:

30

COD Removal, %

25 20 15 10

Reaction time=0.5 h Reaction time=1.0 h Reaction time=1.5 h Reaction time=2.0 h

5 0 0

2

4

6 pH

8

10

12

Fig. 3. Effect of pH on the COD removal. Conditions: T ¼ 1501C; P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 :

35 30 COD Removal, %

The major role played by the pressure is to maintain the wastewater oxidation in the aqueous phase. In addition, high pressure maintained in the reactor also enhances the DO concentration in the liquid phase, which in turns provides a strong driving force for oxidation. To see the effect of pressure on the oxidation reaction, experiments were conducted under two different pressures (0.65 and 0:80 MPa). In both pressures, the experiments were carried out again at a temperature of 1501C with 1:57  105 m3 s1 airflow rate, keeping pH constant at 3.7. Fig. 2 shows the effect of pressure on the COD removal. From the figure it is seen that the effect of pressure is important and COD removal increases with an increase in pressure. For 2:0 h reaction time, COD removal increased from 20.8% to 29.5%, as the pressure is increased from 0.65 to 0:80 MPa: Although the reactor used permits much higher temperature and pressure operations, lower pressure ð0:65 MPaÞ was chosen primarily because of lower associated capital and operating costs of such a WAO process in practice in comparison to those of supercritical system. Several experiments were conducted to determine whether pH has any influence on COD removal. These experiments were performed at a temperature of 1501C; and at a pressure of 0:65 MPa; with an airflow rate of 1:57  105 m3 s1 at two pH’s namely, 7.0 and 10.0, with different reaction times. For pH adjustment, 5 mol dm3 NaOH was used. The results are shown in Fig. 3. In all these experiments, COD removals increased first with rising pH, then by passing through maximums fell again. The effect of pH on the COD removal was however different for different reaction times. For example, for 0:5 h reaction time, the COD removal increased from 3.9% to 19.7% (an increase of 400%) as the pH was increased from 3.7 to 7.0; for a reaction time of 2:0 h; this increase was only from 20.0% to 25.7% (an increase of 29%). As a result of these experiments, it was decided to perform subsequent experiments at pH ¼ 7: One of the most important operating variables of any chemical reaction is the temperature. The effect of this

1173

25 20 15 10

T=140 °C T=150 °C T=160 °C

5 0 0

1800

3600 5400 Reaction time (s)

7200

9000

Fig. 4. Effect of temperature on the COD removal. Conditions: P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 7:0

variable on the COD removal of the wastewater used under a constant pressure of 0:65 MPa with an airflow rate of 1:57  105 m3 s1 and at pH ¼ 7:0 is shown in Fig. 4. It is apparent in this figure that COD removal increases with rising temperature. For example, while only 22.7% COD removal is achieved at 1401C in 2:0 h reaction time, 32.1% COD removal can be achieved at 1601C in 2:0 h reaction time (an increase of 41.0%). The rate of COD removal increases sharply up to 0:5 h reaction time, after that it increases smoothly. Although specific reaction products were not identified in these experiments, it is reasonable to assume that some of the organic compounds degrade easily to intermediates, such as short chain acids [11], which are refractory to total oxidation but are biodegradable. Due to this, COD removal at start is sharp. For practical purposes, it is not wise to employ excessively high temperatures because of high energy consumption and high capital and operating costs. Hence, a temperature of 1501C would be deemed sufficient for the present WAO process.

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

40

40

35

35

30

30

COD Removal, %

COD Removal, %

1174

25 20 15 10

20 15

Cu+2:Mn+2

Ni+2

Fe+2

Cu+2

Co+2

0 Fe+2:Ni+2

0 Without catalyst

5

Catalyst Type

40

30

20

Without catalyst FeCl2 (0.25 kg m-3) FeCl2 (0.50 kg m-3)

10

0 0

1800

3600 5400 Reaction time (s)

7200

9000

Fig. 6. Effect of catalyst loading on the COD removal. Conditions: T ¼ 1501C; P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 7:0:

The effect of catalyst type on COD removal is shown in Fig. 5. From this figure it is seen that approximately the same COD removals for 2:0 h reaction time are achieved with all the catalysts used. It is also obvious that there is no significant advantage of using mixed catalyst on COD removal. For example, while COD removal is 32.8% with CuðNO3 Þ2 catalyst, this is only 34.0% with mixed catalyst of CuðNO3 Þ2 þ MnCl2 (only 4% increase). FeCl2 was chosen as catalyst in subsequent experiments because of cost advantage. Fig. 6 compares the effect of catalyst loading on the COD removal. Catalyst loading results indicate that the presence of FeCl2 increases the COD removal. The

with air with air-catalyst with oxygen with oxygen-catalyst

10

5

Fig. 5. Effect of catalyst type on the COD removal. Conditions: T ¼ 1501C; P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 7:0; reaction time ¼ 2:0 h; catalyst loading ¼ 1 kg m3 :

COD Removal, %

25

0

1800

3600 5400 Reaction time (s)

7200

9000

Fig. 7. Effect of gas source on the COD removal at T ¼ 1501C and pH ¼ 7:0 without catalyst and with FeCl2 ð0:25 kg m3 Þ catalyst.

COD removal for 2:0 h reaction time increased from 25.7% without catalyst to 33.2% with 0:25 kg m3 catalyst. As the concentration of catalyst was increased, the COD removal did not change very much. For example, the COD removal for 2:0 h reaction time increased from 33.2% with 0:25 kg m3 catalyst to 34.3% with 0:50 kg m3 catalyst. This results demonstrate that the effect of catalyst loading on the COD removal is not very important within the range used. Thus, catalyst loading was chosen as 0:25 kg m3 in the later experiments. Both air and oxygen were tested to determine their effects on the COD removal. Based on the literature survey, air and oxygen flow rates were selected as 1:57  105 m3 s1 [7], and 1:25  105 m3 s1 ; respectively. The results are shown in Fig. 7. From this figure it is seen that without catalyst, a COD removal of 25.7% was achieved for 2:0 h reaction time with air. But when oxygen was used instead of air, COD removal was 35.0% (an increase of 36.0%). On the other hand, with catalyst, increase in the COD removal with oxygen instead of air was not so great. This is also obvious from Fig. 7 that while 33.2% COD removal was achieved for 2:0 h reaction time with air, only 36.0% COD removal was achieved with O2 for the same reaction time (8.5% increase). BOD5 =COD ratio is a very important index for any waste to be biodegradable. Normally, an industrial wastewater with a BOD5 =COD ratio of about 0.5 or above is considered as a biodegradable waste. Therefore, in this study, BOD5 =COD ratio of some reactor effluents was determined. Initial BOD5 and COD of our wastewater are 3.95 and 26:65 kg m3 ; respectively. So, the initial BOD5 =COD ratio is about 0.15. This result indicates that our wastewater is not suitable for biological treatment. After the WAO process without catalyst, at a temperature of 1501C; at a pressure of

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

1175

Table 2 Effect of catalysts on BOD5 =COD ratio of the wastewater oxidized, at T ¼ 1501C; P ¼ 0:65 MPa; airflow rate ¼ 1:57  105 m3 s1 ; pH ¼ 7:0 and reaction time ¼ 2:0 h Catalyst type

Catalyst loading ðkg m3 Þ

COD ðkg m3 Þ

BOD5 ðkg m3 Þ

BOD5 =COD ratio

Without catalyst CuðNO3 Þ2 CoðNO3 Þ2 NiðNO3 Þ2 FeCl2 FeCl2 þ NiðNO3 Þ2 CuðNO3 Þ2 þ MnCl2

— 0.25 0.25 0.25 0.25 0.125+0.125 0.125+0.125

19.80 17.90 18.40 17.70 17.80 18.60 17.60

8.00 8.05 11.30 7.05 9.80 11.00 11.25

0.40 0.45 0.61 0.40 0.55 0.59 0.64

0:65 MPa and with an airflow rate of 1:57  105 m3 s1 for 2:0 h reaction time, the BOD5 and COD of the effluent were found as 8.0 and 19:8 kg m3 ; respectively. So, BOD5 =COD ratio is now 0.4. Although improved, this ratio is still not sufficient in many cases for biological treatment. On the other hand, WAO processes, under the same conditions but with catalyst, produced the result for BOD5 =COD ratios above 0.4 depending upon the type of the catalyst. These results are shown in Table 2.

4. Conclusions To consider ‘‘wet air oxidation’’ as a pretreatment method to convert bioresistant organics to more readily biodegradable intermediates, knowledge of the impact of the operating conditions such as temperature, pressure, pH, etc. is required since the overall effectiveness of an integrated chemical–biological process greatly depends on the performance of the chemical pretreatment step. In the light of these, wet air oxidation of alcaloide factory wastewater containing high concentration of various bioresistant organic compounds was investigated. Experimental results indicated that over 26% COD removal of the wastewater could be achieved in 2:0 h of reaction time at 1501C; 0:65 MPa and with an airflow rate of 1:57  105 m3 s1 : The experimental data also revealed that the pressure and temperature effects on the COD removal were important. The COD removal was observed to increase with an increase in both pressure and temperature. Maximum COD removal was obtained at around pH ¼ 7:0: Adding a catalyst to the solution significantly improved the COD removal (about 33%) and BOD5 =COD ratio of the effluent. Two loadings of FeCl2 catalyst as 0.25 and 0:50 kg m3 were tested. It was found that there was no significant advantage of increasing catalyst loading from 0.25 to 0:50 kg m3 : Except NiðNO3 Þ2 and CuðNO3 Þ2 ; all the other catalysts tested were effective to elevate the

BOD5 =COD ratio of the treated wastewater from 0.15 to above 0.50. However, these homogeneous catalysts may lead to a secondary pollution problem such that further treatment is required to remove or recover the metal ions from the water after the organic compounds have been oxidized.

Acknowledgements The authors wish to thank the Government Planning Foundation of Turkey for the financial support under Grant No. 98K120560.

References [1] Mishra VS, Mahajani VV, Joshi JB. Wet air oxidation. Ind Eng Chem Res 1995;34:2–48. [2] Zimmermann FJ, Diddams DG. The Zimmermann process and its application in the pulp and paper industry. TAPPI 1960;43(8):710–5. [3] Joglekar HS, Samant SD, Joshi JB. Kinetics of wet air oxidation of phenol and substituted phenols. Water Res 1991;25(2):135–45. [4] Scott JP, Ollis DF. Integration of chemical and biological oxidation process for water treatment: review and recommendations. Environ Prog 1995;14:88–103. [5] Kawabata N, Urano H. Improvement of biodegradability of organic compounds by wet oxidation. Mem Fac Eng Des, Kyoto Inst Technol Ser Sci Technol 1985;34:64–71. [6] Lin SH, Chuang TS. Wet air oxidation and activated sludge treatment of phenolic wastewater. J Environ Sci Health 1994;A 29(3):547–64. [7] Lin SH, Ho SJ. Treatment of desizing wastewater by wet air oxidation. J Environ Sci Health 1996;A 31(2):355–66. [8] Mantzavinos D, Hellenbrand R, Metcalfe IS, Livingston AG. Partial wet oxidation of p-coumaric acid: oxidation intermediates, reaction pathways and implications for wastewater treatment. Water Res 1996;30(12):2969–76.

1176

Y. Kac-ar et al. / Water Research 37 (2003) 1170–1176

[9] Tagashira Y, Takagi H, Inagaki K. Wet-high pressure wastewater treatment in the presence of copper. Jpn Kokai Tokyo Koho JP 75106862, 1975. [10] Goto S, Levec J, Smith JM. Trickle bed oxidation reactors. Catal Rev Sci Eng 1977;15:187–267. [11] Imamura S, Sakai T, Ikuyama T. Wet oxidation of acetic acid catalyzed by copper salt. Sekiyu Gakkai Shi 1982; 25:74–80.

[12] Standard methods for water and wastewater examination, 17th ed. Washington, DC: Am. Public Health Assn. (APHA), 1992. [13] Holzer K, Horak O, Lawson J. Loprox: a flexible way to pretreat poorly biodegradable effluents. Proceedings of the 46th Industrial Waste Conference, 1991, p. 521–30.