Use of Inexpensive Additives in Pyrolysis of Oil Sludge - Energy

Influence of waste brick kiln dust on pyrolytic conversion of polypropylene in to potential automotive fuels. Imtiaz Ahmad , M. Ismail Khan , Hizbulla...
1 downloads 0 Views 109KB Size
102

Energy & Fuels 2002, 16, 102-108

Use of Inexpensive Additives in Pyrolysis of Oil Sludge Je-Lueng Shie,† Ching-Yuan Chang,*,† Jyh-Ping Lin,‡ Duu-Jong Lee,§ and Chao-Hsiung Wu| Graduate Institute of Environmental Engineering, National Taiwan University, Taipei 106, Taiwan, Department of Environmental Engineering, Lan-Yang College of Technology, Tou-Cheng, I-Lan 261, Taiwan, Department of Chemical Engineering, National Taiwan University, Taipei 106, Taiwan, and Department of Environmental Engineering, Da-Yeh University, Chang-Hwa 515, Taiwan Received April 10, 2001. Revised Manuscript Received September 13, 2001

Previous efforts were made to convert the oil sludge into useful resources such as lower molecule organic compounds and carbonaceous residues by pyrolysis with the carrier gas of N2. The liquid products (condensates of gases at 298 K) obtained from the pyrolysis of oil sludge are close to diesel oil. However, they contain a significant amount of vacuum residues of about 9.57 wt %, which decrease the qualities of liquid products. In the present study, the oil sludge from the oil storage tank of a typical petroleum refinery plant located in the northern Taiwan is used as the raw material for the pyrolysis. The influences of using inexpensive and nonharmful additives on the possible improvement of the pyrolysis of oil sludge are investigated. The additives employed include two groups: (1) aluminum compounds (Al, Al2O3, and AlCl3), and (2) iron compounds (Fe, Fe2O3, FeSO4‚7H2O, FeCl3, and Fe2(SO4)3‚nH2O). For the increases of conversion X, the additives provide the offers on the order of Fe2(SO4)3‚nH2O > Fe2O3 > AlCl3 > FeSO4‚7H2O > Al2O3 > FeCl3 > Al > Fe > no additives. It appears that the above additives enhance the reaction rates r when the temperatures T are in 650-710 K, following the orders AlCl3 > Al > Al2O3 > no additives, and Fe2O3 ∼ Fe2(SO4)3‚nH2O > FeCl3 ∼ Fe ∼ FeSO4‚7H2O > no additives at 710 K. The additives achieve the improvement of the quality q of the oil of pyrolysis (as sum of light and heavy naphtha and light gas oil) on the order of Fe2O3 > Fe2(SO4)3‚nH2O > no additives > Al > FeSO4‚7H2O > Al2O3 > Fe > FeCl3 > AlCl3. Nevertheless, the additives improve the liquid yields Y on the order of Al > Fe2(SO4)3‚nH2O > Fe > Fe2O3 > FeCl3 > no additives > AlCl3 > FeSO4‚7H2O > Al2O3. All this information is useful not only to the improvement of a pyrolysis system but also to the better utilization of liquid oil products.

Introduction Due to the minimal alternative energy sources, like hydraulic, solar, windmill, geothermal, and civilianunacceptable nuclear fission energies, fossil fuels are the main energy sources in Taiwan. For both economic and environmental reasons, the transformation of wastes, such as waste oil,1 petroleum vacuum residues,2-4 oil sludge,5-7 and spent plastics8-10 to effective energy †

Graduate Institute of Environmental Engineering. Department of Environmental Engineering, Lan-Yang College. Department of Chemical Engineering. | Department of Environmental Engineering, Da-Yeh University. (1) Sanjay, H. G.; Tarrer, A. R.; Marks, C. Iron-Based Catalysts for Coal/Waste Oil Coprocessing. Energy Fuels 1994, 8, 99-104. (2) Mochida, I.; Zhao, X. Z.; Sakanishi, K. Supressing of Sludge Formation by Two-Stage Hydrocracking of Vacuum Residue at High Conversion. Ind. Eng. Chem. Res. 1990, 29, 2324-2327. (3) Hajdu, P. E.; Tierney, J. W.; Wender, I. Effect of Catalytic Hydropretreatment of Petroleum Vacuum Resid on Coprocessing with Coal. Energy Fuels 1996, 10, 493-503. (4) Joo, H. K.; Hool, J. N.; Curtis, C. W. Determination of Effective Conditions for Two-Stage Coprocessing of Coal with Waste Plastics and Petroleum Resid. Energy Fuels 1999, 13, 1128-1134. (5) Ayen, R. J.; Swanstrom, C. P. Low-Temperature Thermal Treatment of Petroleum Refinery Waste Sludges. Environ. Prog. 1992, 11 (5), 127-133. (6) Shie, J. L.; Chang, C. Y.; Lin, J. P.; Wu, C. H.; Lee, D. J. Resources Recovery of Oil Sludge by Pyrolysis: Kinetics Study. J. Chem. Technol. Biotechnol. 2000, 75, 1-8. ‡ §

sources has became important recently. Among the crude oil refining processes, a lot of oil sludge accumulate from petroleum refineries. Oil sludge is the main solid waste of a petroleum refinery and contains a large amount of combustibles with high heating values. Therefore, oil sludge is a useful recycling resource and the conversion of oil sludge into fuels has been recognized as an attractive approach. In the previous study, the one-, two- and threereaction models were proposed to predict the pyrolysis experimental results of oil sludge.6 This is reasonable regarding the complex compositions of oil sludge. Extensive study has been conducted on the analyses of major products obtained from the pyrolysis of oil sludge in a separate paper.7 From the results of the pyrolysis (7) Chang, C. Y.; Shie, J. L.; Lin, J. P.; Wu, C. H.; Lee, D. J.; Chang, C. F. Major Products Obtained from the Pyrolysis of Oil Sludge. Energy & Fuels 2000, 14, 1176-1183. (8) Wu, C. H.; Chang, C. Y.; Hor, J. L. On The Thermal Treatment of Plastic Mixtures of MSW: Pyrolysis Kinetics. Waste Manage. 1993, 13, 221-235. (9) Lin, J. P.; Chang, C. Y.; Wu, C. H.; Shih, S. M. Thermal Degradation Kinetics of Polybutadiene Rubber. Polymer Degradation and Stability 1996, 53, 295-300. (10) Maksimova, N. I.; Krivoruchko, O. P. Study of Thermocatalytic decomposition of Polyethylene and Polyvinyl Alcohol in the Presence of an Unsteady-State Fe-containing Catalyst. Chem. Eng. Sci. 1999, 54, 4351-4357.

10.1021/ef0100810 CCC: $22.00 © 2002 American Chemical Society Published on Web 12/15/2001

Pyrolysis of Oil Sludge

of oil sludge, at 873 K, the final product distributions relative to the initial dry oil sludge, in wt %, are about 69.63 liquid oils, 3.57 gaseous products, and 13.1 solid residues, respectively. The distillation characteristics of liquid product from the pyrolysis of oil sludge is close to diesel oil. However, it contains a significant amount of vacuum residues of about 9.57 wt %. This affects the quality of pyrolysis oil and its use. For the improvement of pyrolysis oil, upgrading the pyrolysis oil or adding catalysts or additives in oil sludge to advance the pyrolysis reactions is necessary. Treatment of wastes by catalytic cracking has some advantages over thermal degradation (noncatalytic method). Addition of a useful catalyst or additive in the catalytic cracking process at adequate reaction conditions has a great potential to shorten the cracking time, lower the required temperature, increase the cracking ability of wastes, reduce the proportion of solid residue, and/or narrow the product distribution.11 There is still no data about the catalytic cracking effects on the sludge pyrolysis. However, catalytic crackings of gas oils,12-14 coal/wastes oil,1 Canola oil,15 coal/waste plastics/petroleum residues,4 and Arabian vacuum residues16 have been investigated. Farag et al.14 used the unsteady-state pulse technique with gas oil pulses reacting with a fluid catalytic cracking (FCC) catalyst to study the roles of metal traps in a FCC catalyst contaminated with high levels of nickel and vanadium: 3000 ppm Ni and 4500 ppm V. The catalyst, steamed to achieve equilibrium conditions, was artificially impregnated with an antimony compound (0-2100 ppm) and with nickel and vanadium naphthenates. A four-lump model was employed to describe the experimental results and to obtain the kinetic constants. Experimental data with FCC technique showed that the selectivity to gasoline as well as the gasoline yield was significantly improved, coke formation was reduced, and gas formation was increased.14 Katikaneni et al.15 studied the conversion of Canola oil (used as a representative feed material for waste oils and fats) in the presence and absence of steam using silica-alumina, HZSM-5, and four hybrid catalysts. In most cases, HZSM-5 provided a high selectivity for aromatic hydrocarbons than silicaalumina catalyst, while the selectivity for aliphatic hydrocarbons was higher with silica-alumina than HZSM-5 catalyst in the OLPs (organic liquid products). When zeolite catalysts were added to silica-alumina catalyst, the coke formation and OLP yields decreased whereas the gas yields increased. In general, cracking and aromatization reactions increased significantly with the addition of H-Y or HZSM-5 to silica-alumina.15 (11) Chiu, S. J.; Cheng, W. H. Thermal Degradation and Catalytic Cracking of Poly(ethylene terephthalate). Polym. Degrad. Stab. 1999, 63, 407-412. (12) Larocca, M.; Ng, S.; de Lasa, H. Fast Catalytic Cracking of Heavy Gas Oil: Modeling Coke Deactivation. Ind. Eng. Chem. Res. 1990a, 29, 171-180. (13) Larocca, M.; Farag, H.; Ng, S.; de Lasa, H. Cracking Catalyst Deactivation by Nickel and Vanadium Contaminants. Ind. Eng. Chem. Res. 1990b, 29, 2181-2191. (14) Farag, H.; Ng, S.; de Lasa, H. Kinetic Modeling of Catalytic Cracking of Gas Oils Using in Situ Traps (FCCT) to Prevent Metal Contaminant Effects. Ind. Eng. Chem. Res. 1993, 32, 1071-1080. (15) Katikaneni, S. P.; Adjaye, J. D.; Bakhshi, N. N. Studies on the Catalytic Conversion of Canola Oil to Hydrocarbons: Influence of Hybrid Catalysts and Steam. Energy Fuels 1995, 9, 599-609. (16) Mochida, I.; Zhao, X. Z.; Sakanishi, K. Catalytic Two-Stage Hydrocracking of Arabian Vacuum Residue at a High Conversion Level without Sludge Formation. Ind. Eng. Chem. Res. 1990, 29, 334-337.

Energy & Fuels, Vol. 16, No. 1, 2002 103

However, in coal/waste oil coprocessing, the use of ironbased catalyst precursors and a traditional hydrogen donor solvent such as tetralin with waste oil did not have a significant effect on the conversion and selectivity during coprocessing.1 Nevertheless, the sulfur removal and the ash removal from the waste oil increased.1 In the coprocessing of coal with waste plastics and petroleum reside, Joo et al.4 evaluated four factors, each with three levels, namely, the catalyst type (NiMo/ Al2O3, NiMo/zeolite, and HZSM-5), the residue type (Manji, Maya, and Hondo), the plastic type (low-density polyethylene and mixed postconsumer and mixed waste plastics collected in Germany) and the weight percent of coal (0, 10, and 29 wt %). The most effective conditions to achieve maximum conversion and the production of hexane solubles were the NiMo/zeolite catalyst, Manji residue, and a postconsumer mixture of waste plastics, with no coal being present.4 Mochida et al.16 studied the catalytic two-stage hydrocracking of Arabian vacuum residue by using commercial Ni-Mo catalysts in a batch autoclave to achieve a higher conversion above 50% to 540 °C distillate without producing “dry sludge”, which is defined as insoluble substances in the product oil matrix. The hydrogenation at 390 °C of the first stage was found to be very effective in suppressing sludge formation, and in the second stage at higher temperatures, 430-450 °C, the cracking to produce the distillate dominantly proceeded. The shorter contact time of the second stage at a relatively higher temperature appeared favorable to increase the conversion without producing the sludge.16 Wei and Tsai17 studied the lead immobilization during the pyrolysis of lead-containing oil sludge with the addition of inorganic absorbents of Al2O3 or SiO2. The addition of inorganic absorbents lowered the TCLP (toxicity characteristics leaching procedure) value of lead. Ohtsuka and Asami18 focused on novel methods of converting inexpensive raw (iron and calcium compounds) materials to active catalysts for low-temperature coal gasification in a thermobalance system under atmospheric pressure. Among the precipitation methods using NH3, urea, and Ca(OH)2, the use of Ca(OH)2 provided the most active iron from an aqueous solution of FeCl3, which led to a coal gasification with a 17-fold rate enhancement at 1023 K. Meanwhile, Ca(OH)2 promoted the gasification at 973 K for all the coals examined when kneaded in water. The exchanged catalyst of CaCO3 achieved a 40-60-fold rate enhancement for the gasification at 973 K of brown coal. Thus CaCO3 is the most promising catalyst material for steam gasification of low-rank coals.18 Other predominant and attractive aspects of catalyst usage is in the chemical recycling of waste plastics into the corresponding monomers or raw chemicals that could be reused for the production of plastics or other advanced materials. Solid acids19-21 (such as acidic (17) Wei, Y. L.; Tsai, J. W. Lead Immobilization and Pyrolysis of Lead - containing Oil Sludge. J. Chin. Inst. Environ. Eng. (Taiwan) 2000, 10 (4), 255-260. (18) Ohtsuka, Y.; Asami, K. Highly Active Catalysts from Inexpensive Raw Materials for Coal Gasification. Catal. Today 1997, 39, 111125. (19) Sakata, Y.; Uddin, M. A.; Muto, A. Degradation of Polyethylene and Polypropylene into Fuel Oil by Using Solid Acid and Non-acid Catalysts. J. Anal. Appl. Pyrolysis 1999, 51, 135-155. (20) Park, D. W.; Hwang, E. Y.; Kim, J. R.; Choi, J. K.; Kim, Y. A.; Woo, H. C. Catalytic Degradation of Polyethylene over Solid Acid Catalysts. Polymer Degrad. Stab. 1999, 65, 193-198.

104

Energy & Fuels, Vol. 16, No. 1, 2002

zeolites (HZSM-5, HMOR, and HUSY), non-zeolites (such as SiO2/Al2O3) and silicalite (Si/Al >1000, siliceous analogue of ZSM-5 zeolite, mesoporous silica gel, and mesoporous folded silica)), metal chlorides (CuCl2‚2H2O and MgCl2‚6H2O), metal oxides (Al2O3, Sb2O3, MgO, ZnO), metal acetates (Zn(OAc)2‚2H2O, Sn(OAc)2, Ca(OAc)2‚H2O, Mn(OAc)2‚4H2O),11 metal powders (Al, Zn, Fe, Ni and Cu),22 and ceramic materials (alumina, mullite, silica, mesoporous silica (KFS-16))23,24 have been preferably employed as catalysts for the degradation of waste plastics. The effects and activities of these catalysts depend on the waste plastic types, temperature, contacting time, mixing ways, etc. In the poly(ethylene terephthalate) (PET) thermal degradation and catalytic cracking reactions, copper (II) chloride was most active among the tested catalysts.11 It was noted that some catalysts with highly catalytic activities reduced the liquid yields and increase the gas products. For example, catalysts possessing strong acid sites, such as zeolite ZSM-5, accelerated the degradation of PP (polypropylene) and PE (polyethylene) into gases which resulted in low liquid yields.19 Also, in the determination of degradation kinetics of waste plastics, Day et al.25 pointed out that, in the presence of copper, iron oxide (Fe2O3), and dirt, the metal contamination influenced the degradation behavior of the pure polymers. Applying the first-order kinetics to the Arrhenius parameters suggested that the degradation process was accelerated about 20% for the polymers ABS (acrylonitrile butadiene styrene), PU (polyurethane), and PVC (poly(vinyl chloride)), while a much larger increase of over 100% was noted for the PP sample.25 The previous studies provided some useful results of the catalytic degradation of fossil wastes. However, most of the catalysts used are expensive materials. Although most of them can be recovered from the residues, they would be deactivated by poisoning (such as sulfur26) after several runs. Therefore, commercially viable catalysts/additives must be abundant, inexpensive, and nonharmful, especially when the raw materials used are wastes. For this reason, the present research interest has been directed toward the use of iron and aluminum compounds as catalysts/additives. There is still no data about the catalytic cracking effects on oil sludge pyrolysis. In the present article, the catalytic degradations of oil sludge in the presence of inexpensive and nonharmful additives of (1) aluminum compounds (Al, Al2O3, and AlCl3) and (2) iron compounds (Fe, Fe2O3, FeSO4‚7H2O, FeCl3, and Fe2(SO4)3‚nH2O) are studied in detail. The catalytic degradation is performed by the use of a (21) Lin, Y. H.; Sharratt, P. N. Convertion of Waste Plastics to Hydrocarbons by Catalytic Zeolited Pyrolysis. J. Chinese Inst. Environ. Eng. (Taiwan) 2000, 10 (4), 271-277. (22) Xi, G.; Rui, L.; Tang, Q.; Li, J. Mechanism Studies on the Catalytic Degradation of Waste Polystyrene into Styrene in the Presence of Metal Powders. J. Appl. Polym. Sci. 1999, 73, 1139-1143. (23) Yang, T. C. K.; Chang, W. H.; Viswanath, D. S. Thermal Degradation of Poly(vinyl butyral) in Alumina, Mullite and Silica Composites. J. Thermal Anal. 1996, 47, 697-713. (24) Sakata, Y.; Uddin, M. A.; Muto, A.; Kanada, Y.; Koizumi, K.; Murata, K. Catalytic Degradation of Polyethylene into Fuel Oil over Mesoporous Silica (KFS-16) Catalyst. J. Anal. Appl. Pyrolysis 1997, 43, 15-25. (25) Day, M.; Cooney, J. D.; MacKinnon, M. Degradation of Contaminated Plastics: A Kinetic Study. Polymer Degrad. Stabil. 1995, 48, 341-349. (26) Tomita, A.; Ohtsuka, Y.; Tamai, Y. Nickel Catalyzed Gasification of Brown Coal in a Fluidized Bed Reactor. Fuel 1983, 62, 150153.

Shie et al.

dynamic thermogravimetric (TG) reaction system at the temperature-programmed heating rate (HR) of 5.2 K/min in nitrogen atmosphere of the range of 378-740 K. Experimental Section Materials. The oil sludge used in this study is sampled from the crude oil storage tank of a typical petroleum refinery plant located in the northern Taiwan. Nitrogen gas, used for the purging gas, with 99.99% purity, was purchased from the Ching-Feng-Harng Co. Ltd. in Taipei, Taiwan. The oil sludge sample was dried in a recycle ventilation drier for 24 h at 378 K before use. Experimental Procedure. Thermogravimetry (TG). The laboratory-scale apparatus and detailed description of experimental procedures for the pyrolysis of oil sludge were the same as those in the previous study.6-9 The heating rate employed for the pyrolysis of oil sludge without and with additives was 5.2 K/min. The dried sample was placed on a quartz disk enclosed in a quartz shell and tube reactor with purging nitrogen gas for 2 h at 378 K for further drying to a constant mass prior to starting the temperature rise under a specific heating rate. A sample of known mass (1000 mg of oil sludge or 1000 mg of oil sludge with 100 mg of additives) was used, and the flow rate was adjusted to the desired value, say, 50 cm3/min under 101.3 kPa (1 atm) and 293 K. The effluent gas was cold-trapped at 298 K and then vented to a fume hood. When the run was finished, the nitrogen gas was kept flowing until the temperature of the system was below 373 K. Several duplicate experimental runs were performed in order to collect sufficient amounts of solid residues and liquid oils for analysis. Sampling. The mass of oil sludge used for the experiments of study on the pyrolysis products was 1000 mg or 1000 mg with 100 mg of additives. The products of pyrolysis of oil sludge without or with additives are divided into solid residues, liquid oils (condensable liquid, 298 K), and noncondensable gases (298 K) which were vented to a fume hood. The programmed rising temperature for collecting solid residues was set to stop at 740 K. The condensates collected in 378-740 K were the liquid oils. The connecting glass line between the pyrolysis furnace and condensing tube was wrapped with a thermal belt at 410 K. Analysis. A Hewlett-Packard (HP 5890 series II) gas chromatograph (GC) with an injector port and a flame ionization detector (FID) was used for the quantitative analyses of liquid oils. The chromatographic column was a Supelco fused silica capillary column (SPB-5, 30 m, 0.53 mm i.d., 1.5 µm film thickness). An integrator from Hewlett-Packard (HP 3395) was connected to the GC for graphing and integrating purposes. The operating conditions for the simulation distillation of liquid products and commercial oils were set as follows: injector temperature 393 K, detector temperature 473 K, column temperature following the sampling injection being held at 313 K for 1 min, programmed to 573 K at 10 K/min, and finally held at 573 K for 20 min, nitrogen carrier gas flow rate 3.5 mL/min, nitrogen makeup gas 26 mL/min, sample volume 0.1 µL. The elemental analyses for the solid residues were made on a Perkin-Elmer, Norwalk, CT2400 elemental analyzer with 0.3 wt % accuracy, i.e., C, H, and N analyzed with Heraeus CHN-O-RAPID, and S and Cl analyzed with Tacussel Coulomax 78 automatic coulometric titrator. Chemicals. The principal liquid standards for establishing calibration curve were the same as those in the previous study.7 The liquid standards for the elemental analyzer are sufonilic acid, 1-chloro-2, 4-dinitrobenzene, 3,5-dinitrobenzoic acid, acetanilide, benzoic acid, and stearic acid. Fifteen chemicals are used as additives and directly added (physically mixed) into the quartz disk with oil sludge. They include two groups: (1) aluminum compounds (Al, Al2O3, and AlCl3), and

Pyrolysis of Oil Sludge

Figure 1. Residual mass fractions (M′ ) 1 - X′) without deducting corresponding masses of additives for pyrolysis of oil sludge without and with aluminum compound additives at 5.2 K/min heating rate. (2) iron compounds (Fe, Fe2O3, FeSO4‚7H2O, FeCl3, and Fe2(SO4)3‚nH2O). All of the compounds are commercially available and directly used without further purification. Quantitative analysis of liquid products is based on the calculation using the linear calibration response equations of standards. The equation is generated for each of the liquid standards using a minimum of five different concentrations with three replicates at each concentration. All correlation coefficients r2 of the linear calibration response curves are greater than 0.996.

Results and Discussion Thermal Degradation of Additives. In the sole pyrolysis of oil sludge, the residual mass fraction of oil sludge M decreases as the thermal pyrolysis temperature T increases.6 With the addition of additive into the oil sludge, the mass residual fraction of additive Ma may decrease along with that of oil sludge during pyrolysis. To clarify the actual mass loss of oil sludge only, a variation of residual mass fractions of sample M′ and Ma must be examined and compared. The residual mass fraction of sample (oil sludge and additive) is expressed as

M′ ) W′/W′o where W′ ) m, m ) present mass of sample at T, and W′o ) m o, mo ) initial mass of sample. Figure 1 shows the results of M′ of the absence and presence of aluminum compounds (Al, Al2O3, and AlCl3). The magnitudes of M′ follow the order (1) with Al, (2) with Al2O3, (3) without additives, and (4) with AlCl3. The residual mass fraction of sole additive Ma in thermal degradation at heating rate HR of 5.2 K/min is examined and expressed on a normalized basis, where Ma ) Wa/Wao (Wa, Wao ) present and initial masses of additive, respectively). The results are shown in Figure 2. Under the thermal degradation temperature range of 378-740 K, the values of Ma of Al2O3, Al, and Fe do not have obvious reductions; however, those of FeSO4‚ 7H2O, Fe2(SO4)3‚nH2O, FeCl3, and AlCl3 decrease as T increases. The mass losses of FeSO4‚7H2O and Fe2(SO4)3‚nH2O would be due to the evaporation of water. The oil sludge sample is dried in a recycle ventilation drier for 24 h at 378 K before use. Therefore, the water in the sample of dry oil sludge is negligible. The lost amounts of waters of Fe2(SO4)3‚nH2O and FeSO4‚7H2O are about 16 and 11 wt % of additives, or 1.6 and 1.1 wt % of oil sludge sample, respectively (Figure 2). The pathways of lost waters include (1) evaporation to the

Energy & Fuels, Vol. 16, No. 1, 2002 105

Figure 2. Residual mass fraction (Ma ) 1 - Xa) curves of additives used in this study at 5.2 K/min heating rate.

gas phase, (2) condensation in the liquid products, and (3) reaction with sulfur compounds.29 In practice, some waters are uncollected and coated on the wall of the collection line. For FeCl3 and AlCl3, the acute descending values of Ma are reasonable due to their low evaporating points (with boiling points (bps) of 592 K for FeCl3 and 451 K for AlCl3). Effects of Additives on Conversion of Pyrolysis of Oil Sludge. The oil sludge used in this study is sampled from the crude oil storage tank of a typical petroleum refinery. Its high values of combustible (58.97 wt % of wet basis), heating value of dry basis (10681 kcal/kg), low heating value of wet basis (5870 kcal/kg), and C element (83.94 wt % of dry basis) suggest that the waste of oil sludge would be a valuable resource.6 For the sole pyrolysis of oil sludge, the major gaseous products (noncondensable gases at 298 K) are CO2 (50.88 wt %), HCs (hydrocarbons, 25.23 wt %), H2O (17.78 wt %), and CO (6.11 wt %).7 The HCs mainly consist of low molecular paraffins and olefins (C1-C2, 51.61 wt % of HCs). The distillation characteristics of liquid product (condensate of gas at 298 K) from the sole pyrolysis of oil sludge (collecting temperature range of 378-873 K) is close to diesel oil. However, it contains a significant amount of vacuum residues of about 9.57 wt %.7 For the reduction of vacuum residues, which residues affect the qualities of liquid oils significantly, and lower the reaction temperature, the catalytic degradation of oil sludge in the presence of inexpensive and nonharmful additives is studied. When the mass loss of additive is considered, the residual mass fraction of oil sludge M is expressed as

M ) W/Wo ) 1 - X where W ) m - ma,

Wo ) mo - mao where X ) conversion of oil sludge, m and ma ) present masses of sample and additive at T (mg), and mo and mao) initial masses of sample and additive (mg). Figures 3 and 4 illustrate M vs T with deduction of additives Ma, for the pyrolysis systems with and without (27) Chen, K. C. An Investigation of Major Products for Pyrolysis of Tire Tread in Nitrogen. M. S. Thesis, Graduate Institute of Environmental Engineering, National Taiwan University, Taipei, Taiwan, 1996. (28) Kotanigawa, T.; Yamamoto, M.; Sasaki, M.; Wang, N. Active Site of Iron-Based Catalyst in Coal Liquefaction. Energy Fuels 1997, 11, 190-193. (29) Pradhan, V. R.; Tierney, J. W.; Wender, I. Finely Dispersed Iron, Iron-Molybdenum, and Sulfated Iron Oxides as Catalysts for Coprocessing Reactions. Energy Fuels 1991, 5, 497-507.

106

Energy & Fuels, Vol. 16, No. 1, 2002

Figure 3. Residual mass fractions (M ) 1 - X) for pyrolysis of oil sludge without and with aluminum compound additives at 5.2 K/min heating rate (with deduction of additives Ma).

Figure 4. Residual mass fractions (M ) 1 - X) for pyrolysis of oil sludge without and with iron compound additives at 5.2 K/min heating rate (with deduction of additives Ma).

aluminum and iron compounds, respectively. Two duplicate experimental runs are performed in order to obtain representative reaction data. The standard deviations of the residual mass fractions for the pyrolysis of oil sludge without and with the additive of Fe2(SO4)3‚nH2O are about 0.001-0.018 and 0.001-0.020, respectively. It appears that all the values of M in the presence of additives are lower than those in the absence of additives. Before 650 K, the differences between absence and presence of additives are not obvious. In terms of conversion X () 1 - M), the enhancing effects of additives are on the order of Fe2(SO4)3‚nH2O > Fe2O3 > AlCl3 > FeSO4‚7H2O > Al2O3 > FeCl3 > Al > Fe > no additives. After 650 K, all the conversions with the presence of additives are higher than those without additives. This may be due to the dispersion phenomenon. Sanjay et al.1 pointed out that the dispersant additives present in the waste oil are believed to enhance coal conversion. Further, the results indicate that the degradation of oil sludge with additives can be carried out at a lower temperature to achieve the same degradation (conversion) than that without additives. In other words, it can therefore save the energy and treatment cost for the oil sludge pyrolysis. Comparison of the results with the additives of metal powders (Al and Fe) and without additives (Figures 3 and 4) indicates that the conversions of oil sludge are XAl > XFe > Xno additive at 740 K. The catalytic effects increase with the increasing activities of metals.22 While the electronegativities of metals (Al of 1.61, Fe of 1.81) increase gradually, the forces between the metals and radicals in the transient intermediates also increase gradually. That is to say, the stabilities of the transient intermediates increase as the electronegativities of metals increase.22 Hence, the stable transient inter-

Shie et al.

Figure 5. Instantaneous reaction rate (r ) dX/dt) vs T at 5.2 K/min heating rate for pyrolysis of oil sludge without and with aluminum compound additives (with deduction of additives).

Figure 6. Instantaneous reaction rate (r ) dX/dt) vs T at 5.2 K/min heating rate for pyrolysis of oil sludge without and with iron compound additives (with deduction of additives).

Figure 7. Ratios of instantaneous reaction rates with to that without additives for pyrolysis of oil sludge at 5.2 K/min heating rate (with deductions of additives).

mediates result in the low conversion and activity. This agrees with the experimental results. Effects of Additives on Reaction Rates of Pyrolysis of Oil Sludge. The variations of instantaneous reaction rates (r ) dX/dt) in the presence and absence of additives with the deductions of additives in the pyrolysis of oil sludge are shown in Figures 5 and 6 for Al and Fe compounds, respectively. Their ratios Rpa are also illustrated in Figure 7. For T < 650 K, the difference of reaction rates between presence and absence of additives are negligible except that of Fe2O3. In 650-710 K, the additives enhance the reaction rates, following the orders AlCl3 > Al > Al2O3 > no additives and Fe2O3 ∼ Fe2(SO4)3‚nH2O > FeCl3 ∼ Fe ∼ FeSO4‚ 7H2O > no additives at 710 K, with the values of Rpa of about 1.13-1.23. However, above 720 K, the rates without additives are predominant and larger, that is with Rpa < 1. The more active the additive is, the more tendency toward the lower temperature corresponding to the maximum value of reaction rate is. From these

Pyrolysis of Oil Sludge

Energy & Fuels, Vol. 16, No. 1, 2002 107

Table 1. Elemental Analysis of Oil Sludge and Solid Residues of Pyrolysis of Oil Sludge without and with Additives in This Studyg,h C

H

a initial dry oil sludge without additives with Fe with Fe2(SO4)3‚nH2O with FeCl3 with FeSO4‚7H2O with Fe2O3 with Al with Al2O3 with AlCl3

83.9 (0.90)d 64.39 (0.6) 32.04 (4.19) 34.59 (2.25) 34.76 (1.56) 35.15 (1.87) 36.48 (0.68) 30.55 (0.12) 30.55 (1.42) 41.48 (0.28)

b

a

51.03 62.99 49.84 63.31 66.55 50.24 52.46 60.22

12.0 (0.58) 2.59 (0.06) 1.87 (0.07) 1.39 (0.07) 3 (0.2) 1.49 (0.03) 1.34 (0.03) 1.47 (0.1) 1.67 (0.01) 2.43 (0.16)

N b

a

2.98 2.55 4.3 2.68 2.44 2.42 2.87 3.53

0.81 (0.53) 1 (0.04) 0.5 (0.07) 0.71 (0.06) 0.52 (0.05) 0.57 (0.03) 0.54 (0.02) 0.45 (0.04) 0.52 (0.04) 0.65 (0.01)

b

C/Hc

mass ratio

0.8 1.29 0.75 1.02 0.99 0.74 0.89 0.94

6.99 24.86 17.12 24.7 11.59 23.62 27.27 20.76 18.28 17.05

100 16.34e 23.94 16.8 18.54 18.19 18.48 22.96 21.33 17.15

16.34f 16.15 10.35 14.98 13.39 12.13 15.39 14.04 13.28

a Based on mass of residues (including additives), units in wt % for C, H, and N. b Based on mass of oil sludge residues (with deduction of additives), units in wt % for C, H, and N. c In w/w for C/H ratio (b column). d Values in parentheses are standard deviations (σn-1). e Mass ratio of residues (including additives) to sum of initial dry oil sludge and initial additives added M′. f Mass ratio of oil sludge residues (with deduction of additives) to initial dry oil sludge M. g Heating rate (HR) ) 5.2 K/min. h Final temperature ) 740 K.

results, the favorable reaction temperature range with the presence of additives is 650-710 K. Influences of Additives on Solid Residues. At the final pyrolysis temperature Tf (740 K), the residual mass fractions of sample M′ and oil sludge M are shown in Table 1. Also included are the results of the elemental analysis of solid residues. At 740 K, the lowest M is that with the additive of Fe2(SO4)3‚nH2O. Therefore, it is the most active additive in 378-740 K. Meanwhile, it indicates that carbon (C) is the major element in solid residues. The variations of carbon element of solid residues with the additives of Fe2(SO4)3‚nH2O, Fe2O3, FeSO4‚7H2O, and AlCl3 relative to such variations without additives are not obvious. However, there are sharp reductions of carbon element for the cases with FeCl3, Al, Fe, and Al2O3. It is noted that the C/H values are lower for FeCl3 of 11.59, AlCl3 of 17.05, Fe of 17.12, and Al2O3 of 18.28 than the C/H value without additives of 24.86. This might be due to the low reductions of hydrogen (H) with additives of FeCl3, AlCl3, Fe, and Al2O3. Influences of Additives on Yields and Qualities of Liquid Oils. The liquid oils are collected by passing the gaseous products through a glass connecting line wrapped with a heating tape of 410 K before collecting at 298 K. They are mostly collected in the first condensing tube immersed in 298 K water bath. The liquid yields without and with additives are expressed as

Y(liquid yield, wt %) ) mL/Wo where mL ) total mass of collected liquids (mg),

Wo ) mo - mao (mg), and mo and mao ) initial masses of sample and additives (mg). The liquid yields and solid residues of the pyrolysis of oil sludge without and with additives in 378-740 K are shown in Figure 8. From Figure 8, the liquid yields are 71.03 and 70.75 wt % with the addition of Al and Fe2(SO4)3‚nH2O, respectively. The liquid oils and commercial oils are analyzed for the different boiling points by GC, according to the Standard Test Method for Boiling Range Distribution of Petroleum Fractions, proposed by the ASTM D-2887 method. Five portions of oils are cut apart as listed in Table 2.

Figure 8. Liquid yields and solid residues of pyrolysis of oil sludge without and with additives at 5.2 K/min heating rate in 378-740 K.

The simulated distillation results are shown in Figure 9. From Table 2 and Figure 9, the liquid oil collected in 378-740 K for the pyrolysis of oil sludge without additive contains, in wt %, about 7.45 heavy naphtha, 43.9 light gas oil, 48.3 heavy gas oil, and 0.34 vacuum residues, respectively. This is very close to fuel oil. That collected in 378-873 K contains, in wt %, about 0.72 light naphtha, 12.28 heavy naphtha, 66.58 light gas oil, 10.85 heavy gas oil, and 9.57 vacuum residues, respectively.7 It is obvious that the liquid oil collected in 378740 K has much lower vacuum residues than that collected in 378-873 K. This might be due to the reason that in the lower temperatures, the large molecular weight compounds would not easily evaporate. Also from Table 2 and Figure 9, the additives of Fe2O3 and Fe2(SO4)3‚nH2O improve the qualities (q, in terms of sum of light and heavy naphtha and light gas oil) of liquid oils with Fe2O3 better than Fe2(SO4)3‚nH2O. Nevertheless, the additives of FeCl3 and AlCl3 vigorously decrease the qualities of pyrolysis oils. The other additives (such as Al2O3, Al powder, FeSO4‚7H2O, and Fe powder) do not have obvious effects on the pyrolysis oils. Referring again to Table 1, the C/H ratios of solid residues for the pyrolysis cases with additives of FeCl3 (11.59), and AlCl3 (17.05) are lower than that without additive (24.86). It appears that the metal chlorides inhibit the release of hydrogen thus lower the qualities of liquid oils. Ferric-Based Additives for Pyrolysis of Oil Sludge. For the above additives tested, the most active additive with greatest conversion is Fe2(SO4)3‚nH2O.

108

Energy & Fuels, Vol. 16, No. 1, 2002

Shie et al.

Table 2. Distillation Characteristics of Pyrolysis Oil without and with Additives in this study and some commercial oilsa,b,c

without additives with Fe with Fe2(SO4)3‚nH2O with FeCl3 with FeSO4‚7H2O with Fe2O3 with Al with Al2O3 with AlCl3 GL7 DL7 FL27 HL27

light naphtha 343-366 K

heavy naphtha 366-477 K

light gas oil 477-616 K

heavy gas oil 616-811 K

vacuum residue >811 K

∼0 ∼0 ∼0 ∼0 ∼0 ∼0 ∼0 ∼0 ∼0 10.07 0.04 0 0

7.45d 2.38 7.91 1.8 4.74 15.08 6.37 9.38 0.93 62.93 7.84 4.03 0.29

43.9 41.8 48.86 34.86 43.21 42.8 42.3 37.11 15.19 26.15 87.27 46.01 23.88

48.3 55.82 42.95 62.87 51.62 42.12 50.58 53.19 83.88 0.85 4.78 49.59 75.19

0.34 ∼0 0.29 0.47 0.42 ∼0 0.76 0.32 ∼0 ∼0 ∼0 0.38 0.64

a GL, DL, FL, and HL: gasoline, diesel oil, fuel oil, and heavy oil. 5.2 K/min. d Unit: in wt %.

b

pyrolysis temperature range ) 378-740 K. c Heating rate (HR) )

can hydrogenate the solvents to increase the donatability and also quench the radicals formed by the thermal degradation of coal at an elevated temperature.28 It is believed that these phenomena in coal liquefaction also happen in the pyrolysis of oil sludge with additives of Fe2(SO4)3‚nH2O and Fe2O3. Conclusions

Figure 9. Simulated distillation results of pyrolysis oil without and with additives and some commercial oils.

The additives of Fe2O3 and Fe2(SO4)3‚nH2O improve the liquid qualities (q, in terms of sum of light and heavy naphtha and light gas oil) of pyrolysis oils. Nevertheless, the additives of Al and Fe2(SO4)3‚nH2O give the greatest liquid yields. All these results point out that ferric-based additives (such as Fe2(SO4)3‚nH2O and Fe2O3) are the most favorable additives. The reason may be due to the cause that the oil sludge and Fe2(SO4)3‚nH2O contain about 2.06 wt % (in dry basis) elemental sulfur6 and sulfate, respectively. The additive of Fe2O3 may react with sulfur of oil sludge during pyrolysis, to form FeS2, which is then oxidized to convert the surface of sulfides to sulfate species.28 It was postulated that the sulfate group inhibits the agglomeration of metal oxides and subsequently increases the surface area and catalyst dispersion.29 Kotanigawa et al. 28 also pointed out that the Fe-sulfate catalyst was sufficiently able to catalyze the coal liquefaction without the addition of sulfur, although the presence of elemental sulfur in the catalyst was found to further increase the conversion and selectivity. It seems that the formation of sulfate species on catalyst is essentially important for the coal liquefaction. The sulfate species formed on the surface of iron sulfides can activate hydrogen molecules in the coal liquefaction system. The activated hydrogen molecules

The catalytic degradation is performed using a dynamic thermogravimetric (TG) reaction system at the temperature-programmed heating rate of 5.2 K/min in nitrogen atmosphere in 378-740 K. The influences of using inexpensive and nonharmful additives on the pyrolysis of oil sludge are investigated in this study. The additives employed include two groups: (1) aluminum compounds (Al, Al2O3, and AlCl3), and 2) iron compounds (Fe, Fe2O3, FeSO4‚7H2O, FeCl3, and Fe2(SO4)3‚ nH2O). For the above additives tested, the most active additive with greatest conversion is Fe2(SO4)3‚nH2O. Below 650 K, the differences of reaction rates r between absence and presence of additives are not obvious except that of Fe2O3. It appears that the above additives enhance the values of r in 650-710 K. The more active the additive is, the greater the tendency toward the lower temperature corresponding to the maximum value of reaction rate is. However, above 720 K, the ratios of reaction rates in the presence to absence of additives Rpa are less than 1. There are sharp reductions of carbon element for the cases with FeCl3, Al, Fe, and Al2O3. The additives of Fe2O3 and Fe2(SO4)3‚nH2O improve the liquid qualities (q, in terms of sum of light and heavy naphtha and light gas oil) of pyrolysis oils. Nevertheless, the additives of Al and Fe2(SO4)3‚nH2O give the greatest liquid yields. All this information is useful not only to the proper design of a pyrolysis system but also to the better utilization of liquid oil products. Acknowledgment. We express our sincere thanks to the National Science Council of Taiwan for the financial support, under Contract NSC89-2211-E002-013. EF0100810