- Email: [email protected]
Color and chlorinated organics removal from pulp mills wastewater using activated petroleum coke
PII: S0043-1354(00)00322-5
Wat. Res. Vol. 35, No. 3, pp. 745–749, 2001 # 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0043-1354/01/$ - see front matter
COLOR AND CHLORINATED ORGANICS REMOVAL FROM PULP MILLS WASTEWATER USING ACTIVATED PETROLEUM COKE AYMAN R. SHAWWA*M, DANIEL W. SMITH and DAVID C. SEGO Department of Civil and Environmental Engineering , 304 Environmental Engineering Building, University of Alberta, Edmonton, AB, Canada T6G 2M8 (First received 1 December 1999; accepted in revised form 1 May 2000) Abstract}Delayed petroleum coke, a waste by-product from the oil sand industry, was utilized in the production of activated carbon. The activated carbon was then evaluated for color and chlorinated organics reduction from pulp mill wastewater. The activation of the petroleum coke was evaluated using a fixed bed reactor involving carbonization and activation steps at temperature of 8508C and using steam as the activation medium. The activation results showed that the maximum surface area of the activated coke was achieved at an activation period of 4 h. The maximum surface area occurred at burnoff and water efficiency of 48.5 and 54.3%, respectively. Increasing the activation period to 6 h resulted in a decrease in the surface area. Methylene blue adsorption results indicated that the activation process was successful. Methylene blue adsorbed per 100 g of applied activated coke was 10 times higher than that adsorbed by raw petroleum coke. Adsorption equilibrium results of the bleached wastewater and the activated coke showed that significant color, COD, DOC and AOX removal (>90%) was achieved when the activated coke dose exceeded 15,000 mg/L. Adsorption isotherms, in terms of COD, DOC, UV and color were developed based on the batch equilibrium data. Based on these isotherms, the amount of activated coke required to achieve certain removal of color and AOX can be predicted. The utilization of the petroleum coke for the production of activated carbon can provide an excellent disposal option for the oil sand industry at the same time would provide a cheap and valuable activated carbon. # 2001 Elsevier Science Ltd. All rights reserved Key words}AOX, adsorption, color, isotherm, pulp mills, petroleum coke
NOMENCLATURE
B Ce K M 1=n q R Weff X
burnoff of petroleum coke, % equilibrium concentration of color, mg color/L Freundlich parameter, mg color/mg AC the mass of powdered activated carbon, g Freundlich parameter, dimensionless the amount of color adsorbed per mg of applied activated carbon, mg color/mg AC mass recovery of carbon, % water efficiency, % the adsorbed amount of color, mg
INTRODUCTION
Bleached kraft pulp mill wastewater constitutes one of the most problematic parts of the entire pulp industry. These effluents contain high concentrations of color causing compounds, mainly due to lignin, measured by absorption at 465 nm. Also, when *Author to whom all correspondence should be addressed. Tel.: +1 780-438-7879; fax: +1 780-435-8086; e-mail: [email protected] 745
chlorine is used for bleaching a high concentration of chlorinated organic compounds is present in the wastewater. Discharge of these compounds into a receiving water body can impact the ecological balance and cause aesthetic concerns. In addition, chlorinated organic compounds may be associated with some toxicity and they show considerable resistance to biological and chemical degradation. All pulp mills in Alberta use some sort of biological treatment systems, which are successful in the reduction of biological oxygen demand (BOD) and total suspended solids (TSS) concentration in their effluent. However, biological treatment is not very effective for reducing color and chlorinated organics in pulp mill wastewater (C¸ec¸en et al., 1992). There are other effective treatments alternatives for color reduction in bleached kraft mill effluent. Activated carbon adsorption is considered very effective in the reduction of color, adsorbable organic halides (AOX) and the nonbiodegradable fraction of the pulp bleaching wastewater. However, this process has costs associated with the production of activated carbon.
746
Ayman R. Shawwa et al.
One option for substantially reducing the high cost associated with using activated carbon adsorption is by producing activated carbon from an abundant and cheap source of raw carbonaceous materials. An abundant and cheap source in many regions including western Canada is petroleum coke, which is a byproduct of the upgrading of bitumen from oil sands into synthetic crude oil. This petroleum coke can be utilized as a cheap and readily available raw material for the production of more valuable activated carbon suitable for the removal of color and AOX from pulp bleaching wastewater. Previous studies on the activation of petroleum coke showed that the activated coke have adsorptive capacities 10 time greater than the raw coke (Shawwa et al., 1999). The adsorptive capacity was attributed to the large surface area and porosity developed from the activation process. The objectives of this study are: (i) to evaluate the production of powdered activated carbon from petroleum coke using fixed bed reactor involving carbonization and activation steps and using steam as the activation media, and (ii) to evaluate the activated powdered coke for color and AOX reduction from bleached pulp mill wastewater.
EXPERIMENTAL
In this study, delayed petroleum coke from Suncor Canada was used as the source of raw petroleum coke. Delayed petroleum coke is produced in a coking drum heated to 4808C. Volatile material is removed and the remaining by-product is broken up hydraulically and then removed from the drum. The delayed coker particles are irregularly shaped and has an amorphous structure, where 92% of the carbon atoms are aromatic (Majid et al., 1989). The chemical composition and the ash content (in terms of w/w%) of delayed coke are given in Table 1. The chemical characteristics and ash content of the 5-590-A coconut charcoal, which was used in this study for comparative purposes, are also shown in Table 1. Delayed coke samples were dried in a vacuum oven for 24 h at a temperature of 1108C prior to grinding. Samples were pulverized using a ball mill and the size range between 200 mesh (44 mm) and 325 mesh (75 mm) was utilized for the activation experiments. The production of the activated coke involved two stages: the carbonization stage and the
activation stage. The carbonization stage was carried out at 850 28C at an average heating rate of 108C/min. The activation stage was carried out immediately after the carbonization using steam. Activation periods ranging from 1 to 6 h were tested. All carbonization and activation experiments were carried out in a horizontal quartz reactor with a diameter of 45 mm, as shown in Fig. 1. The reactor was placed horizontally in a tube furnace with an accurate temperature controller and under N2 atmosphere. The carbonization and activation procedures involved placing 10 g of raw petroleum coke sample in the reactor. A metal thermocouple was inserted in one side of the reactor such that its tip is just above the sample. Water was introduced to the reactor by a syringe pump as a steam source for activation. It was deoxygenated before use by boiling. The water was introduced into the reactor at a rate of 0.5 mL/ min. A detailed experimental procedure for the activation of petroleum coke is presented in Shawwa et al. (1999). The success of activation of the delayed petroleum coke was evaluated by performing liquid-phase adsorption using methylene blue dye as adsorbate. Sorption capacities for methylene blue would provide a valuable indication of the capacity of the activated carbon to remove color from wastewater (Sontheimer et al., 1988). The adsorption procedure followed the ASTM D3860-98 (ASTM, 1999). Batch adsorption tests were carried out by contacting the activated coke samples with pulp mill bleaching effluent in order to generate adsorption isotherms. The characteristics of the pulp mill bleaching effluent, which was obtained from Weyerhaeuser Canada Ltd., Alberta, Canada, are given in Table 2. For this purpose, 250 mL sealed flasks were filled with 100 mL of the bleaching effluent and different amounts of activated coke samples were added. The mixture in each flask was mixed using a tumbler for 24 h. It was assumed that the equilibrium was achieved in 24 h. Subsequently, the wastewater was separated from the activated coke by pressure filtration through 0.45 mm cellulose nitrate filters. The filtrate was analyzed, according to standard methods (APHA, AWWA and WEF, 1992), for color measured at wavelength of 465 nm (color465), ultra violet absorption at wavelength of 254 nm (UV254), chemical oxygen demand (COD), dissolved organic carbon (DOC) and AOX.
Table 1. Chemical composition and ash content of delayed petroleum coke and coconut charcoal (after Watkinson et al., 1989 and Majid et al., 1989) Parameter
Volatile (wt%) Moisture content (wt%) BET-surface area (m2/g) Iodine number Ash content (wt%) SiO2 (wt%) Al2O3 (wt%) Fe (wt%) Ti (wt%) Ni (wt%) V (wt%)
Delayed petroleum coke
5-590-A coconut charcoal
11.9 1.8 9 28 3 41.7 19.2 16 1.5 1.6 2.8
} 2 1237 1110 7 } } } } } }
Fig. 1. Experimental setup for the activation of petroleum coke.
Table 2. Pulp mill bleaching effluent characteristics Parameter COD (mg/L) DOC (mg/L) AOX (mg/L as Cl) UV254 nm (cm ÿ 1) Color465 nm (mg/L Pt–Co) pH
Concentration 2126 575 80.2 13.11 2300 2.1
Color and chlorinated organics removal from pulp mills wastewater RESULTS AND DISCUSSIONS
The activation of delayed petroleum coke consisted of seven samples, in duplicate, activated with steam for activation periods ranging from 0 to 6 h. The activation time of 0 h corresponded to petroleum coke samples that were subjected to carbonization only and no activation with steam. The mass recovery (R) and burnoff (B) of petroleum coke samples were calculated for each activation period. The mass recovery (R) is defined as the weight of the activated coke divided by the original weight of the sample multiplied by 100%. On the other hand, B is the mass of carbon that was oxidized as a result of the activation with steam and is defined as 100% R. In addition, the amount of water that reacted with petroleum coke, referred to as the water efficiency (Weff ), was also calculated. The Weff is defined as the mass of water, which produced steam and reacted with the coke sample over the mass of the total water used in the activation. The activation results showed that the burnoff of petroleum coke (B) increased from 8 to 67% as the activation time increased from 0 to 6 h. The burnoff of coke resulted from the reaction between steam and tar deposits inside the initial pore structure of the coke sample. As the activation time proceeded, the steam widens the pore structure further and increasing the surface area. The activation process is carried out until the maximum surface area is achieved in which 50% of the carbon structure has been burnt out (Jankowska et al., 1991). This can be illustrated in Fig. 2, which shows the effect of activation time on Weff and B. The results in Fig. 2 indicated that there was a steady rise in Weff going from 0 to a maximum value of 54.3% after 4 h of activation. The Weff values then declined slowly to approximately 40% after 6 h of activation. These results suggested that there were few surface sites for carbon–steam reaction to occur at the start of the activation because of the low surface area, as indicated by low B values. As the activation time proceeded, more surface area was exposed and thus more active sites for carbon–steam reaction, as indicated by high Weff and B values. The
Fig. 2. Water efficiency of activation as a function of burnoff.
747
maximum Weff value of 54.3% occurred when the maximum surface area was achieved (at B value of 48.5%). Beyond this point of activation, further oxidation starts to burn the walls of the pores decreasing the surface area and therefore Weff values decreased. The success of the activation of delayed petroleum coke was evaluated by performing liquid-phase adsorption using methylene blue as adsorbate. The adsorption capacities for raw petroleum coke, activated petroleum coke and 5-590-A coconut charcoal are summarized in Fig. 3. The adsorptive capacity of raw petroleum coke was 8 mg/g of applied coke. The methylene blue adsorption ranged from 10 to 100 mg/g for coke samples activated from 0 to 6 h. The maximum adsorption occurred for coke samples activated for 4 h with methylene blue adsorption of 100.5 mg/g of applied activated coke. The methylene blue adsorption for the coconut charcoal had much higher adsorption capacity of 420 mg/g of applied activated carbon due to its large surface area (BET value of 1237 m2/g). Color removal of the bleaching effluent was evaluated at different activated carbon dosages using the 2, 4 and 6 h activated coke samples. Figure 4 shows color465 concentration at activated coke dosages ranging from 100 to 15,000 mg/L. The results showed that the 4 h activated coke sample achieved the highest color removal as the dose increased. This
Fig. 3. Methylene blue adsorption results for different activation periods.
Fig. 4. Color removal at different activated coke dose and activation periods.
748
Ayman R. Shawwa et al.
was expected since the maximum surface area of activated coke was achieved after 4 h of activation. In the dosage range of 100 to 2500 mg/L, the highest removal of color was about 33%. As the activated coke dose increased from 2500 to 15,000 mg/L, color removal increased to 90%. Similar trends were observed for COD, DOC, UV254 and AOX. In addition to the decrease in pollutant concentrations, it was noted that activated coke addition to the bleached effluent has changed its composition, as shown in the COD/DOC and color465/DOC versus activated coke dose diagram, which is illustrated in Fig. 5. The decrease in the specific oxygen demand (COD/DOC) with increasing activated coke dose indicated that the remaining solution contained more oxidized compounds than the untreated bleached wastewater (Sontheimer et al., 1988). The COD/ DOC ratio decreased significantly when the activated coke dose was higher than 2500 mg/L. The activated coke removed the recalcitrant portion of the organic compounds present in the bleached wastewater. As a result, the remaining solution in the bleached wastewater would become more susceptible to biological treatment (Yin et al., 1990). In addition, the results in Fig. 5 showed that significant color465 removal was achieved when the activated coke dose was higher than 2500 mg/L. The high color removal exceeded that achieved in COD or DOC, which indicates the preferential removal of high molecular weight chlorolignins from bleached wastewater (C¸ec¸en, 1993; Ying and Tucker, 1990). Activated coke addition also led to a significant decrease in AOX concentration. The AOX, or absorbable organic halide, is considered one of the commonly used surrogate measures for chlorinated organics concentration in wastewater. It measures volatile halides that are associated with organic compounds (Gergov et al., 1988). For activated coke dose ranging from 100 to 2500 mg/L, the highest removal of AOX was 26% for coke samples activated for 4 h. Increasing the carbon dose from 2500 to 15,000 mg/L resulted in 90% reduction in AOX. Adsorptive capacity for activated petroleum coke was estimated from adsorption isotherm data for color465. The Freundlich isotherm is widely used for a
variety of heterogeneous adsorption systems because it gives more accurate results than the Langmuir isotherm (Chen and Horan, 1998). Therefore, the adsorption data were fitted by least-squares regression into the Freundlich equation of the form
Fig. 5. COD/DOC and Color/DOC ratios at different activated coke dose (4 h-activation period).
Fig. 6. Adsorption isotherm for activated coke with 4 hactivation time.
q ¼ X=M ¼ KCen where q is the amount of color465 adsorbed per mg of applied activated coke, X is adsorbed amount of color465, M is the weight of carbon, Ce is the equilibrium concentration of color465 in solution and K and n are Freundlich adsorption constants. The Freundlich constant K represents the adsorption capacity of the carbon for specific adsorbate, at a given equilibrium concentration Ce . In general, a high value for K is desirable for higher color reduction. Adsorption isotherms for color465, COD, DOC and UV254 were developed for activated coke with 4 h activation period. The isotherm data for color465 were plotted as log q versus log Ce shown in Fig. 6. The bleached effluent contained both easily and poorly adsorbable compounds, which competed for activated carbon sites. At activated coke dosages ranging from 100 to 2500 mg/L, only the easily adsorbable compounds were removed from the solution and the equilibrium concentration is determined by the poorly adsorbable compounds left in the solution (Sontheimer et al., 1988). As a result, a steep isotherm is obtained, which is shown in region 1 of Fig. 6. Similar trends were observed for COD, DOC and UV254. The Freundlich constants K and n based on region 1 were determined as 5 10 ÿ 22 mg color/mg activated coke and 6.6, respectively. The adsorption isotherm for color465 also showed that as the activated coke dose increased from 2500 to 15000 mg/L, the amount of easily adsorbable compounds per unit mass decreased. The isotherm flattened and would reflect only the behavior of poorly adsorbable compounds, or the high molecular weight chlorolignins (Chen and Horan, 1998). This part of the isotherm is represented by region 2 of Fig. 6. Similar trends were observed for COD, DOC and UV254. The Freundlich constants K and n based on region 2 were determined as 0.05 mg color/mg activated coke and 0.2, respectively.
Color and chlorinated organics removal from pulp mills wastewater Table 3. Adsorption constants based on region 2 of the adsorption isotherma Parameter COD DOC Color465 nm UV254 nm a
K
n
0.0457 0.0241 0.0514 0.0093
0.2303 0.2541 0.2024 0.4611
The K values for color, COD and DOC are given in mg/mg activated coke and for UV in cm ÿ 1/mg activated coke.
The Freundlich constants K and n based on region 2 were used to calculate the adsorption capacity for the activated coke because true adsorption of chlorolignins can be followed in this region. The adsorption capacity for bleaching effluent that has color of 2300 mg/L is 0.25 mg color/mg activated coke. The adsorption constants K and n expressed as color465 and UV254 were of the same order of magnitude as these found in terms of COD and DOC, as shown in Table 3. In general, better adsorptivity was obtained when the adsorption was expressed in terms of color465 than DOC and COD. The UV analysis is reported to reflect the removal of halogenated organics better than DOC (Sontheimer et al., 1988). Therefore, the rapid UV analysis can be substituted for DOC and COD analysis when investigating the adsorption of AOX compounds from bleaching wastewater. CONCLUSIONS
The results from this study confirmed that the activation process of petroleum coke was successful. The maximum surface area of the activated coke was developed after 4 h of activation with steam at 8508C. This was confirmed by the methylene blue adsorption results which showed that the highest adsorption occurred using the 4 h-activated coke and that the adsorption capacity was 10 times higher than that for the raw petroleum coke. The adsorption equilibrium results showed that significant color and AOX removal (>90%) from the bleaching wastewater was achieved when activated coke dose exceeded 15,000 mg/L. The decrease in the specific oxygen demand (COD/DOC) with increasing activated coke dose indicated that the remaining solution in the bleaching wastewater contained more oxidized compounds and as a result it would be more susceptible to biological treatment. Therefore, adsorption of color and chlorinated organics with activated coke can be utilized as a pretreatment step before any
749
biological treatment. Based on the adsorption isotherms developed in this study, the amount of activated coke required to achieve certain removal of color and chlorinated organics can be predicted. Acknowledgements}This project was funded in part by the Sustainable Forest Management Network of Centers of Excellence and part by a research grant from the Natural Sciences and Engineering Research Council of Canada. The authors would like to acknowledged Suncore Canada for providing the petroleum coke samples and Weyerhaeuser Canada for providing the bleached stage wastewater samples used in this study.
REFERENCES
APHA, AWWA and WEF (1992) Standard Methods for the Examination of Water and Wastewater, 18th eds., A. Eaton, E. Clesceri and A. Greenberg. Washington, DC. ASTM (1999) Standard Practice for Determination of Adsorptive Capacity of Activated Carbon by Aqueous Phase Isotherm Technique. American Society for Testing and Materials, ASTM D3860-98, West Conshohocken, PA. C¸ec¸en F. (1993) Adsorption characteristics of a biotreated pulp mill effluent. Water Sci. Technol. 28(2), 1–10. C¸ec¸en F., Urban W. and Haberl R. (1992) Biological and advanced treatment of sulfate pulp bleaching effluents. Water Sci. Technol. 26(1–2), 435–444. Chen W. and Horan N. (1998) The treatment of high strength pulp and paper mill effluent for wastewater reuse. Part III: tertiary treatment options for pulp and paper mill wastewater to achieve effluent recycle. Environm. Technol. 19, 173–182. Gergov M., Priha M., Talka E., Valttila O., Kangas A., Kukkonen K. (1988) Chlorinated organic compounds in effluent treatment at kraft mills. Tappi J. 75, 12, 175–184. Jankowska H., Swiatkoski A. and Choma J. (1991) In Activate Carbon, ed T. J. Kemp. Ellis Horwood Limited, New York. Majid A., Ratcliffe C. I. and Ripmeester J. A. (1989) Demineralization of petroleum cokes and fly ash samples obtained from the upgrading of athabasca oil sands bitumen. Fuel Sci. Technol. Int. 7(5–6), 879–895. Shawwa A., Smith D. and Sego D. (1999) Color and Chlorinated Organic Reduction in Kraft Pulp Mill Wastewater Using Activated Petroleum Coke. Sustainable Forest Management Network of Centers of Excellence, Report MIT-6, Edmonton, Canada. Sontheimer H., Crittenden J., Summers R., Fettig J., Horner G., Hubele C. and Zimmer G. (1988) Activated Carbon for Water Treatment. AWWA Research Foundation, Denver, CO. Watkinson A. P., Cheng G. and Func D. P. C. (1989) Gasification of oil sand coke. Fuel 68(1), 4–10. Yin C., Joyce T. and Change H. (1990) Characterization and biological treatment of bleach plant effluent. Proceedings of the 44th Purdue Industrial Waste Conference. Lewis Publishers, Michigan. Ying W. and Tucker M. (1990) Selecting activated carbon for adsorption treatment. Proceeding of the 44th Purdue Industrial Waste Conference, Lewis Publishers, Michigan.
Wat. Res. Vol. 35, No. 3, pp. 745–749, 2001 # 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0043-1354/01/$ - see front matter
COLOR AND CHLORINATED ORGANICS REMOVAL FROM PULP MILLS WASTEWATER USING ACTIVATED PETROLEUM COKE AYMAN R. SHAWWA*M, DANIEL W. SMITH and DAVID C. SEGO Department of Civil and Environmental Engineering , 304 Environmental Engineering Building, University of Alberta, Edmonton, AB, Canada T6G 2M8 (First received 1 December 1999; accepted in revised form 1 May 2000) Abstract}Delayed petroleum coke, a waste by-product from the oil sand industry, was utilized in the production of activated carbon. The activated carbon was then evaluated for color and chlorinated organics reduction from pulp mill wastewater. The activation of the petroleum coke was evaluated using a fixed bed reactor involving carbonization and activation steps at temperature of 8508C and using steam as the activation medium. The activation results showed that the maximum surface area of the activated coke was achieved at an activation period of 4 h. The maximum surface area occurred at burnoff and water efficiency of 48.5 and 54.3%, respectively. Increasing the activation period to 6 h resulted in a decrease in the surface area. Methylene blue adsorption results indicated that the activation process was successful. Methylene blue adsorbed per 100 g of applied activated coke was 10 times higher than that adsorbed by raw petroleum coke. Adsorption equilibrium results of the bleached wastewater and the activated coke showed that significant color, COD, DOC and AOX removal (>90%) was achieved when the activated coke dose exceeded 15,000 mg/L. Adsorption isotherms, in terms of COD, DOC, UV and color were developed based on the batch equilibrium data. Based on these isotherms, the amount of activated coke required to achieve certain removal of color and AOX can be predicted. The utilization of the petroleum coke for the production of activated carbon can provide an excellent disposal option for the oil sand industry at the same time would provide a cheap and valuable activated carbon. # 2001 Elsevier Science Ltd. All rights reserved Key words}AOX, adsorption, color, isotherm, pulp mills, petroleum coke
NOMENCLATURE
B Ce K M 1=n q R Weff X
burnoff of petroleum coke, % equilibrium concentration of color, mg color/L Freundlich parameter, mg color/mg AC the mass of powdered activated carbon, g Freundlich parameter, dimensionless the amount of color adsorbed per mg of applied activated carbon, mg color/mg AC mass recovery of carbon, % water efficiency, % the adsorbed amount of color, mg
INTRODUCTION
Bleached kraft pulp mill wastewater constitutes one of the most problematic parts of the entire pulp industry. These effluents contain high concentrations of color causing compounds, mainly due to lignin, measured by absorption at 465 nm. Also, when *Author to whom all correspondence should be addressed. Tel.: +1 780-438-7879; fax: +1 780-435-8086; e-mail: [email protected] 745
chlorine is used for bleaching a high concentration of chlorinated organic compounds is present in the wastewater. Discharge of these compounds into a receiving water body can impact the ecological balance and cause aesthetic concerns. In addition, chlorinated organic compounds may be associated with some toxicity and they show considerable resistance to biological and chemical degradation. All pulp mills in Alberta use some sort of biological treatment systems, which are successful in the reduction of biological oxygen demand (BOD) and total suspended solids (TSS) concentration in their effluent. However, biological treatment is not very effective for reducing color and chlorinated organics in pulp mill wastewater (C¸ec¸en et al., 1992). There are other effective treatments alternatives for color reduction in bleached kraft mill effluent. Activated carbon adsorption is considered very effective in the reduction of color, adsorbable organic halides (AOX) and the nonbiodegradable fraction of the pulp bleaching wastewater. However, this process has costs associated with the production of activated carbon.
746
Ayman R. Shawwa et al.
One option for substantially reducing the high cost associated with using activated carbon adsorption is by producing activated carbon from an abundant and cheap source of raw carbonaceous materials. An abundant and cheap source in many regions including western Canada is petroleum coke, which is a byproduct of the upgrading of bitumen from oil sands into synthetic crude oil. This petroleum coke can be utilized as a cheap and readily available raw material for the production of more valuable activated carbon suitable for the removal of color and AOX from pulp bleaching wastewater. Previous studies on the activation of petroleum coke showed that the activated coke have adsorptive capacities 10 time greater than the raw coke (Shawwa et al., 1999). The adsorptive capacity was attributed to the large surface area and porosity developed from the activation process. The objectives of this study are: (i) to evaluate the production of powdered activated carbon from petroleum coke using fixed bed reactor involving carbonization and activation steps and using steam as the activation media, and (ii) to evaluate the activated powdered coke for color and AOX reduction from bleached pulp mill wastewater.
EXPERIMENTAL
In this study, delayed petroleum coke from Suncor Canada was used as the source of raw petroleum coke. Delayed petroleum coke is produced in a coking drum heated to 4808C. Volatile material is removed and the remaining by-product is broken up hydraulically and then removed from the drum. The delayed coker particles are irregularly shaped and has an amorphous structure, where 92% of the carbon atoms are aromatic (Majid et al., 1989). The chemical composition and the ash content (in terms of w/w%) of delayed coke are given in Table 1. The chemical characteristics and ash content of the 5-590-A coconut charcoal, which was used in this study for comparative purposes, are also shown in Table 1. Delayed coke samples were dried in a vacuum oven for 24 h at a temperature of 1108C prior to grinding. Samples were pulverized using a ball mill and the size range between 200 mesh (44 mm) and 325 mesh (75 mm) was utilized for the activation experiments. The production of the activated coke involved two stages: the carbonization stage and the
activation stage. The carbonization stage was carried out at 850 28C at an average heating rate of 108C/min. The activation stage was carried out immediately after the carbonization using steam. Activation periods ranging from 1 to 6 h were tested. All carbonization and activation experiments were carried out in a horizontal quartz reactor with a diameter of 45 mm, as shown in Fig. 1. The reactor was placed horizontally in a tube furnace with an accurate temperature controller and under N2 atmosphere. The carbonization and activation procedures involved placing 10 g of raw petroleum coke sample in the reactor. A metal thermocouple was inserted in one side of the reactor such that its tip is just above the sample. Water was introduced to the reactor by a syringe pump as a steam source for activation. It was deoxygenated before use by boiling. The water was introduced into the reactor at a rate of 0.5 mL/ min. A detailed experimental procedure for the activation of petroleum coke is presented in Shawwa et al. (1999). The success of activation of the delayed petroleum coke was evaluated by performing liquid-phase adsorption using methylene blue dye as adsorbate. Sorption capacities for methylene blue would provide a valuable indication of the capacity of the activated carbon to remove color from wastewater (Sontheimer et al., 1988). The adsorption procedure followed the ASTM D3860-98 (ASTM, 1999). Batch adsorption tests were carried out by contacting the activated coke samples with pulp mill bleaching effluent in order to generate adsorption isotherms. The characteristics of the pulp mill bleaching effluent, which was obtained from Weyerhaeuser Canada Ltd., Alberta, Canada, are given in Table 2. For this purpose, 250 mL sealed flasks were filled with 100 mL of the bleaching effluent and different amounts of activated coke samples were added. The mixture in each flask was mixed using a tumbler for 24 h. It was assumed that the equilibrium was achieved in 24 h. Subsequently, the wastewater was separated from the activated coke by pressure filtration through 0.45 mm cellulose nitrate filters. The filtrate was analyzed, according to standard methods (APHA, AWWA and WEF, 1992), for color measured at wavelength of 465 nm (color465), ultra violet absorption at wavelength of 254 nm (UV254), chemical oxygen demand (COD), dissolved organic carbon (DOC) and AOX.
Table 1. Chemical composition and ash content of delayed petroleum coke and coconut charcoal (after Watkinson et al., 1989 and Majid et al., 1989) Parameter
Volatile (wt%) Moisture content (wt%) BET-surface area (m2/g) Iodine number Ash content (wt%) SiO2 (wt%) Al2O3 (wt%) Fe (wt%) Ti (wt%) Ni (wt%) V (wt%)
Delayed petroleum coke
5-590-A coconut charcoal
11.9 1.8 9 28 3 41.7 19.2 16 1.5 1.6 2.8
} 2 1237 1110 7 } } } } } }
Fig. 1. Experimental setup for the activation of petroleum coke.
Table 2. Pulp mill bleaching effluent characteristics Parameter COD (mg/L) DOC (mg/L) AOX (mg/L as Cl) UV254 nm (cm ÿ 1) Color465 nm (mg/L Pt–Co) pH
Concentration 2126 575 80.2 13.11 2300 2.1
Color and chlorinated organics removal from pulp mills wastewater RESULTS AND DISCUSSIONS
The activation of delayed petroleum coke consisted of seven samples, in duplicate, activated with steam for activation periods ranging from 0 to 6 h. The activation time of 0 h corresponded to petroleum coke samples that were subjected to carbonization only and no activation with steam. The mass recovery (R) and burnoff (B) of petroleum coke samples were calculated for each activation period. The mass recovery (R) is defined as the weight of the activated coke divided by the original weight of the sample multiplied by 100%. On the other hand, B is the mass of carbon that was oxidized as a result of the activation with steam and is defined as 100% R. In addition, the amount of water that reacted with petroleum coke, referred to as the water efficiency (Weff ), was also calculated. The Weff is defined as the mass of water, which produced steam and reacted with the coke sample over the mass of the total water used in the activation. The activation results showed that the burnoff of petroleum coke (B) increased from 8 to 67% as the activation time increased from 0 to 6 h. The burnoff of coke resulted from the reaction between steam and tar deposits inside the initial pore structure of the coke sample. As the activation time proceeded, the steam widens the pore structure further and increasing the surface area. The activation process is carried out until the maximum surface area is achieved in which 50% of the carbon structure has been burnt out (Jankowska et al., 1991). This can be illustrated in Fig. 2, which shows the effect of activation time on Weff and B. The results in Fig. 2 indicated that there was a steady rise in Weff going from 0 to a maximum value of 54.3% after 4 h of activation. The Weff values then declined slowly to approximately 40% after 6 h of activation. These results suggested that there were few surface sites for carbon–steam reaction to occur at the start of the activation because of the low surface area, as indicated by low B values. As the activation time proceeded, more surface area was exposed and thus more active sites for carbon–steam reaction, as indicated by high Weff and B values. The
Fig. 2. Water efficiency of activation as a function of burnoff.
747
maximum Weff value of 54.3% occurred when the maximum surface area was achieved (at B value of 48.5%). Beyond this point of activation, further oxidation starts to burn the walls of the pores decreasing the surface area and therefore Weff values decreased. The success of the activation of delayed petroleum coke was evaluated by performing liquid-phase adsorption using methylene blue as adsorbate. The adsorption capacities for raw petroleum coke, activated petroleum coke and 5-590-A coconut charcoal are summarized in Fig. 3. The adsorptive capacity of raw petroleum coke was 8 mg/g of applied coke. The methylene blue adsorption ranged from 10 to 100 mg/g for coke samples activated from 0 to 6 h. The maximum adsorption occurred for coke samples activated for 4 h with methylene blue adsorption of 100.5 mg/g of applied activated coke. The methylene blue adsorption for the coconut charcoal had much higher adsorption capacity of 420 mg/g of applied activated carbon due to its large surface area (BET value of 1237 m2/g). Color removal of the bleaching effluent was evaluated at different activated carbon dosages using the 2, 4 and 6 h activated coke samples. Figure 4 shows color465 concentration at activated coke dosages ranging from 100 to 15,000 mg/L. The results showed that the 4 h activated coke sample achieved the highest color removal as the dose increased. This
Fig. 3. Methylene blue adsorption results for different activation periods.
Fig. 4. Color removal at different activated coke dose and activation periods.
748
Ayman R. Shawwa et al.
was expected since the maximum surface area of activated coke was achieved after 4 h of activation. In the dosage range of 100 to 2500 mg/L, the highest removal of color was about 33%. As the activated coke dose increased from 2500 to 15,000 mg/L, color removal increased to 90%. Similar trends were observed for COD, DOC, UV254 and AOX. In addition to the decrease in pollutant concentrations, it was noted that activated coke addition to the bleached effluent has changed its composition, as shown in the COD/DOC and color465/DOC versus activated coke dose diagram, which is illustrated in Fig. 5. The decrease in the specific oxygen demand (COD/DOC) with increasing activated coke dose indicated that the remaining solution contained more oxidized compounds than the untreated bleached wastewater (Sontheimer et al., 1988). The COD/ DOC ratio decreased significantly when the activated coke dose was higher than 2500 mg/L. The activated coke removed the recalcitrant portion of the organic compounds present in the bleached wastewater. As a result, the remaining solution in the bleached wastewater would become more susceptible to biological treatment (Yin et al., 1990). In addition, the results in Fig. 5 showed that significant color465 removal was achieved when the activated coke dose was higher than 2500 mg/L. The high color removal exceeded that achieved in COD or DOC, which indicates the preferential removal of high molecular weight chlorolignins from bleached wastewater (C¸ec¸en, 1993; Ying and Tucker, 1990). Activated coke addition also led to a significant decrease in AOX concentration. The AOX, or absorbable organic halide, is considered one of the commonly used surrogate measures for chlorinated organics concentration in wastewater. It measures volatile halides that are associated with organic compounds (Gergov et al., 1988). For activated coke dose ranging from 100 to 2500 mg/L, the highest removal of AOX was 26% for coke samples activated for 4 h. Increasing the carbon dose from 2500 to 15,000 mg/L resulted in 90% reduction in AOX. Adsorptive capacity for activated petroleum coke was estimated from adsorption isotherm data for color465. The Freundlich isotherm is widely used for a
variety of heterogeneous adsorption systems because it gives more accurate results than the Langmuir isotherm (Chen and Horan, 1998). Therefore, the adsorption data were fitted by least-squares regression into the Freundlich equation of the form
Fig. 5. COD/DOC and Color/DOC ratios at different activated coke dose (4 h-activation period).
Fig. 6. Adsorption isotherm for activated coke with 4 hactivation time.
q ¼ X=M ¼ KCen where q is the amount of color465 adsorbed per mg of applied activated coke, X is adsorbed amount of color465, M is the weight of carbon, Ce is the equilibrium concentration of color465 in solution and K and n are Freundlich adsorption constants. The Freundlich constant K represents the adsorption capacity of the carbon for specific adsorbate, at a given equilibrium concentration Ce . In general, a high value for K is desirable for higher color reduction. Adsorption isotherms for color465, COD, DOC and UV254 were developed for activated coke with 4 h activation period. The isotherm data for color465 were plotted as log q versus log Ce shown in Fig. 6. The bleached effluent contained both easily and poorly adsorbable compounds, which competed for activated carbon sites. At activated coke dosages ranging from 100 to 2500 mg/L, only the easily adsorbable compounds were removed from the solution and the equilibrium concentration is determined by the poorly adsorbable compounds left in the solution (Sontheimer et al., 1988). As a result, a steep isotherm is obtained, which is shown in region 1 of Fig. 6. Similar trends were observed for COD, DOC and UV254. The Freundlich constants K and n based on region 1 were determined as 5 10 ÿ 22 mg color/mg activated coke and 6.6, respectively. The adsorption isotherm for color465 also showed that as the activated coke dose increased from 2500 to 15000 mg/L, the amount of easily adsorbable compounds per unit mass decreased. The isotherm flattened and would reflect only the behavior of poorly adsorbable compounds, or the high molecular weight chlorolignins (Chen and Horan, 1998). This part of the isotherm is represented by region 2 of Fig. 6. Similar trends were observed for COD, DOC and UV254. The Freundlich constants K and n based on region 2 were determined as 0.05 mg color/mg activated coke and 0.2, respectively.
Color and chlorinated organics removal from pulp mills wastewater Table 3. Adsorption constants based on region 2 of the adsorption isotherma Parameter COD DOC Color465 nm UV254 nm a
K
n
0.0457 0.0241 0.0514 0.0093
0.2303 0.2541 0.2024 0.4611
The K values for color, COD and DOC are given in mg/mg activated coke and for UV in cm ÿ 1/mg activated coke.
The Freundlich constants K and n based on region 2 were used to calculate the adsorption capacity for the activated coke because true adsorption of chlorolignins can be followed in this region. The adsorption capacity for bleaching effluent that has color of 2300 mg/L is 0.25 mg color/mg activated coke. The adsorption constants K and n expressed as color465 and UV254 were of the same order of magnitude as these found in terms of COD and DOC, as shown in Table 3. In general, better adsorptivity was obtained when the adsorption was expressed in terms of color465 than DOC and COD. The UV analysis is reported to reflect the removal of halogenated organics better than DOC (Sontheimer et al., 1988). Therefore, the rapid UV analysis can be substituted for DOC and COD analysis when investigating the adsorption of AOX compounds from bleaching wastewater. CONCLUSIONS
The results from this study confirmed that the activation process of petroleum coke was successful. The maximum surface area of the activated coke was developed after 4 h of activation with steam at 8508C. This was confirmed by the methylene blue adsorption results which showed that the highest adsorption occurred using the 4 h-activated coke and that the adsorption capacity was 10 times higher than that for the raw petroleum coke. The adsorption equilibrium results showed that significant color and AOX removal (>90%) from the bleaching wastewater was achieved when activated coke dose exceeded 15,000 mg/L. The decrease in the specific oxygen demand (COD/DOC) with increasing activated coke dose indicated that the remaining solution in the bleaching wastewater contained more oxidized compounds and as a result it would be more susceptible to biological treatment. Therefore, adsorption of color and chlorinated organics with activated coke can be utilized as a pretreatment step before any
749
biological treatment. Based on the adsorption isotherms developed in this study, the amount of activated coke required to achieve certain removal of color and chlorinated organics can be predicted. Acknowledgements}This project was funded in part by the Sustainable Forest Management Network of Centers of Excellence and part by a research grant from the Natural Sciences and Engineering Research Council of Canada. The authors would like to acknowledged Suncore Canada for providing the petroleum coke samples and Weyerhaeuser Canada for providing the bleached stage wastewater samples used in this study.
REFERENCES
APHA, AWWA and WEF (1992) Standard Methods for the Examination of Water and Wastewater, 18th eds., A. Eaton, E. Clesceri and A. Greenberg. Washington, DC. ASTM (1999) Standard Practice for Determination of Adsorptive Capacity of Activated Carbon by Aqueous Phase Isotherm Technique. American Society for Testing and Materials, ASTM D3860-98, West Conshohocken, PA. C¸ec¸en F. (1993) Adsorption characteristics of a biotreated pulp mill effluent. Water Sci. Technol. 28(2), 1–10. C¸ec¸en F., Urban W. and Haberl R. (1992) Biological and advanced treatment of sulfate pulp bleaching effluents. Water Sci. Technol. 26(1–2), 435–444. Chen W. and Horan N. (1998) The treatment of high strength pulp and paper mill effluent for wastewater reuse. Part III: tertiary treatment options for pulp and paper mill wastewater to achieve effluent recycle. Environm. Technol. 19, 173–182. Gergov M., Priha M., Talka E., Valttila O., Kangas A., Kukkonen K. (1988) Chlorinated organic compounds in effluent treatment at kraft mills. Tappi J. 75, 12, 175–184. Jankowska H., Swiatkoski A. and Choma J. (1991) In Activate Carbon, ed T. J. Kemp. Ellis Horwood Limited, New York. Majid A., Ratcliffe C. I. and Ripmeester J. A. (1989) Demineralization of petroleum cokes and fly ash samples obtained from the upgrading of athabasca oil sands bitumen. Fuel Sci. Technol. Int. 7(5–6), 879–895. Shawwa A., Smith D. and Sego D. (1999) Color and Chlorinated Organic Reduction in Kraft Pulp Mill Wastewater Using Activated Petroleum Coke. Sustainable Forest Management Network of Centers of Excellence, Report MIT-6, Edmonton, Canada. Sontheimer H., Crittenden J., Summers R., Fettig J., Horner G., Hubele C. and Zimmer G. (1988) Activated Carbon for Water Treatment. AWWA Research Foundation, Denver, CO. Watkinson A. P., Cheng G. and Func D. P. C. (1989) Gasification of oil sand coke. Fuel 68(1), 4–10. Yin C., Joyce T. and Change H. (1990) Characterization and biological treatment of bleach plant effluent. Proceedings of the 44th Purdue Industrial Waste Conference. Lewis Publishers, Michigan. Ying W. and Tucker M. (1990) Selecting activated carbon for adsorption treatment. Proceeding of the 44th Purdue Industrial Waste Conference, Lewis Publishers, Michigan.