Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales

G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Contents lists available at ScienceDirect Journal of Analytical and Applied Pyrolysis journal homepage: www.elsevier.com/locate/jaap Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales H. da Silva Almeida b,c , O.A. Corrêa a , J.G. Eid a , H.J. Ribeiro a,c , D.A.R. de Castro a,c , M.S. Pereira c,f , L.M. Pereira b,c , A. de Andrade Aâncio a,c , M.C. Santos a,c , S.A.P da Mota d , J.A. da Silva Souza a,c , Luiz E.P. Borges f , N.M. Mendonça b , N.T. Machado a,b,e,∗ a Laboratory of Separation Processes and Applied Thermodynamic (TERM@), Faculty of Chemical Engineering-UFPA, Rua Augusto Corrêia No. 1, CEP: 66075-900, Cx. P. 8619, Belém, Pará, Brazil b Laboratory of Multiuser Water and Sludge Analysis and Treatment (LAMAG), Faculty of Environment and Sanitary Engineering-UFPA, Rua Augusto Corrêia No. 1, CEP: 66075-900, Cx. P. 8619, Belém, Pará, Brazil c Graduate Program of Natural Resource Engineering-UFPA, Rua Augusto Corrêia No. 1, CEP: 66075-900, Cx. P. 8619, Belém, Pará, Brazil d Faculty of Materials Engineering-UNIFESSPA, Quadra 17, Bloco 4, Lote Especial, Nova Marabá, CEP: 68505-080, Marabá, Pará, Brazil e Leibniz-Institüt für Agrartechnik Potsdam-Bornin e.V, Department of Postharvest Technology, Max-Eyth-Allee 100, Potsdam 14469, Germany f Laboratory of Catalyst Preparation and Catalytic Cracking, Section of Chemical Engineering-IME, Praça General Tibúrcio No. 80, CEP: 22290-270 Rio de Janeiro, RJ, Brazil a r t i c l e i n f o Article history: Received 15 February 2016 Received in revised form 23 April 2016 Accepted 30 April 2016 Available online xxx Keywords: Catalytic cracking Grease traps Distillation Green kerosene Performance Production scales a b s t r a c t This work aims to investigate the effect of catalytic cracking of residual fat, oils, and grease (FOG) from grease traps in different production scales (bench, laboratory, and pilot) on the reaction products yields and OLP properties and the feasibility to produce kerosene-like hydrocarbons. The cracking experiments were carried out in batch mode at 450 ◦ C and 1.0 atmosphere, with 10% (wt.) Na2 CO3 using a laboratory scale cylindrical borosilicate-glass reactor of 143 mL, a bench scale stirred tank slurry reactor of 1.5 L, and a pilot scale stirred tank slurry reactor of 143 L (≈1:10:1000). The reaction liquid products were physical and chemical analyzed for acid and saponification values, density, kinematic viscosity, refractive index, and copper strip corrosion. FT-IR analysis provided the qualitative chemical composition of OLP obtained in bench, laboratory, and pilot scales, as well as kerosene, light and heavy diesel-like hydrocarbons fractions obtained by distillation of OLP produced in pilot scale with 10% (wt.) Na2 CO3 . The chemical compositions of OLP and kerosene-like hydrocarbons fraction obtained in pilot scale determined by NMR and GC–MS. The results showed an OLP yield ranging from 62.90 to 66.57% (wt.), a coke yield ranging between 7.02 and 9.79% (wt.), and a gas yield ranging from 16.32 to 22.40% (wt.), showing a mean absolute percentage deviation of 2.12%, 11.88%, and 14.91% for OLP, gas, and coke yields respectively, obtained in different production scales (≈10:1000). The OLP acid values varied from 19.08 to 10.45 mg KOH/g, the density between 0.820 and 0.835 g/cm3 , and the kinematic viscosity from 3.28 to 4.21 mm2 s−1 . The yield of kerosene-like hydrocarbons fraction average 14.90% (wt.) with an acid value of 5.43 mg KOH/g, density of 0.740 g/cm3 , and kinematic viscosity of 0.66 mm2 s−1 , while those of light and heavy diesel-like hydrocarbons fractions average 32.01% (wt.) and 19.35% (wt.) respectively. FT-IR and NMR analysis of OLP and kerosene-like hydrocarbons fraction confirms the presence of functional groups characteristic of hydrocarbons (alkenes, alkanes, ring-containing alkenes, and ring-containing alkanes, and cycloalkanes) and oxygenates (carboxylic acids, ketones, fatty alcohols, and dienes). The GC–MS analysis of OLP and kerosene-like hydrocarbons fraction obtained in pilot scale with 10% (wt.) Na2 CO3 identified in OLP 76.97% hydrocarbons (39.44% alkenes, 31.91% alkanes, 4.12% ring-containing alkenes, and 1.50% ringcontaining alkenes) and 23.03% oxygenates (12.14% carboxylic acids, 6.98% ketones, 1.90% fatty alcohols, and 2.01% dienes). ∗ Corresponding author at: Laboratory of Separation Processes and Applied Thermodynamic (TERM@), Faculty of Chemical Engineering-UFPA, Rua Augusto Corrêia No. 1, CEP: 66075-900, Cx. P. 8619, Belém, Pará, Brazil. E-mail address: machado@ufpa.br (N.T. Machado). http://dx.doi.org/10.1016/j.jaap.2016.04.017 0165-2370/© 2016 Elsevier B.V. All rights reserved. Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 2 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx The kerosene-like hydrocarbons fraction is composed by 94.62% (area) hydrocarbon (44.99% alkenes, 29.61% alkanes, 7.58% ring-containing alkenes, 6.15% ring-containing alkanes, 4.31% cycloalkanes, and 1.98% aromatics) and 5.38% (area) oxygenates (5.38% carboxylic acids), showing that catalytic cracking of scum from grease traps with 10% (wt.) Na2 CO3 is technically feasible. © 2016 Elsevier B.V. All rights reserved. 1. Introduction Food services (restaurants, fast-food restaurants, industrial restaurants, etc.) generates wastewaters streams containing fats, oils, and greases (FOG) and suspended solids (SS) [1–7]. The discharge of fats, oils, and grease-containing wastewaters into public sewers may causes not only a reduction of sanitary sewers performance and flowing capacity [5,7], but also clogging in drain pipes and sewer lines, because of formation of hard deposits [2,7]. In addition, corrosion of sewer lines due to anaerobic reactions may occur [2], posing a waste management challenge for environmentalpublic polices, particularly in developing countries [2,5,7]. The most common pre-treatment process for fats, oils, and greases removal from grease-containing wastewater, prior to discharge into public sewers, is a grease trap [5,7]. However, other processes and/or devices for removal of fats, oils, and greases from grease-containing wastewater streams have been reported in the literature [8,7], including grease trap filters [1,5], ultrafiltration processes [9], and the use of microorganism [10]. In this context, processes have been proposed for the treatment of fats, oils, and grease-containing wastewaters streams, including aerobic removal of FOG by microorganism [10], production of biodiesel [4,6,11–14], and the anaerobic co-digestion of FOG to improve biogas production from anaerobic digesters at wastewater treatment units [7,15,16]. The residual fat material from grease trap, a lipid base material of low quality, consists of fatty acids, frying oils (soybean, sunflower, etc.), animal fats, hydrogenated fats, fatty alcohols, and other compounds [4,8,14,17]. A process that makes it possible the use of lipid base materials of low quality for producing liquid and gaseous fuels is pyrolysis [18–26], and/or catalytic-cracking [17,27–43]. Conversion of low quality lipid materials to produce renewable liquid and gaseous fuels can be either achieved by pyrolysis such as soybean cake [18], safflower seed press cake [19], olive oil residue [20], used sunflower oil [21], waste fish oil [22,23], waste frying oil [24,25], and until industrial fatty wastes (soybean soap stock, beef tallow, and poultry waste) [26], or by thermal-catalytic cracking including fat, oils, and grease (FOG) [17], frying oils [27], used sunflower oil [28], used palm oil and palm oil-based fatty acid mixture [29–34], used vegetable oil [35], fatty acids and animals fat [36–40], animal fat and meat and bone meal [41], and until residues of rendering plants [42]. Most studies on the pyrolysis and catalytic cracking of lipid base material have been carry out in micro [29–33,35,43–48], laboratory [18–21,26–28,34,39,40,49], bench scales [23,26,42,50], and semi-pilot scales [22,24,25,51], a only a few in pilot and/or technical scale [17,36–38,41,52]. Ooi et al. [29–33] studied the catalytic conversion of palm oilbased fatty acid mixture, fatty acids mixture, palm oil fatty acids, and used palm oil, using a fixed bed micro scale reactor (ID = 10 mm, L = 155 mm, VR = 12.17 mL), over HZSM-5 at 400 and 450 ◦ C, 1.0 atm, fatty acid-to-catalyst ratios between 6 and 10, and WHSV between 2.5 and 4.5 h−1 , obtaining at 450 ◦ C and 2.5 h−1 , and OLP yield of 55.8% (wt.) and 40.9% (wt.) gasoline-like fraction [29]; composite MCM-41/␤ zeolite and a mixture of zeolite ␤ and MCM-41 [30], mesoporous materials (AlMCM-41 and LPMM-41) at 450 ◦ C, 1.0 atm, WHSV of 2.5 h−1 , obtaining OLP, gas, coke, and water yields between 48.5 and 63.1%, 6.7 and 25.7%, 1.4 and 11.7%, and 3.2 and 8.8% (wt.), respectively, and gasoline fractions up to 43.0% (wt.), obtained with LPMM-41 [31], over composite catalysts (HZSM-5 and MCM-41/ZSM-5) at 400, 425 and 450 ◦ C, 1.0 atm, and WHSV between 2.5 and 4.5 h−1 , obtaining gasoline fractions up to 52 and 43.0% (wt.) for HZSM-5 and MCM-41/ZSM-5, respectively [32], over composite catalysts composed of microporous HZSM-5 and mesoporous MCM-41/SBA-15 molecular sieve at 450 ◦ C and weight hourly space velocity of 2.5 h−1 , obtaining conversion up 98% (wt.) and gasoline yield 44% (wt.) with HZSM-5 composite catalyst [33]. Ooi Yean Sang [43] investigated the catalytic-cracking of palm oil at 450 ◦ C and 1.0 atm, with HZSM-5 as catalyst, using the apparatus described elsewhere [29], obtaining a OLP composed by hydrocarbons (gasoline, kerosene, and diesel) with a gasoline yield of 48% (wt.). Twaiq et al. [44–47] investigated the catalytic cracking of palm oil, using the apparatus described elsewhere [29], using different catalysts such as MCM-41 with different Si/Al ratios, at 450 ◦ C and 1.0 atm, obtaining conversions between 80 and 90% (wt.), with high selectivity to liquid hydrocarbons [44], zeolites (HZSM-5, K−HZSM-5, zeolite ␤, and USY), at temperatures between 350 and 450 ◦ C and 1.0 atm, producing mainly OLP, gas, and water, with conversions up to 99% (wt.) and gasoline yield of 28% (wt.) at 350 ◦ C [45], composite zeolite ZSM-5/mesoporous molecular sieve, at 450 ◦ C and 1.0 atm, obtaining conversions of 80–100% (wt.) and gasoline yields between 38 and 47% (wt.) [46], MCM41, at 450 ◦ C and 1.0 atm, obtaining linear hydrocarbons (C13 ). The yield of OLP decreased with increasing specific catalyst surface area, and the gasoline selectivity increased, while that of diesel decreased with increasing conversion of palm oil [47]. Siswanto et al. [48] investigated the catalytic cracking of palm oil to produce gasoline, at 450 ◦ C and 1.0 atm, using a fixed bed micro reactor (150 mm × 25 mm ID, VR = 73.63 mL), over MCM-41, and oil-tocatalyst ratios between 30 and 50, obtaining an OLP yields up to 60.73% (wt.) and gasoline yield up to 43.63% (wt.). The yield of OLP decreases with increasing oil/catalyst ratio. Witchakorn Charusiri and Tharapong Vitidsant [35], studied the conversion of used vegetable oil into liquid fuels, at the temperature interval of 400–430 ◦ C, reaction time between 30 and 90 min, and H2 pressure between 10 and 30 bar, over sulfated zirconia, using a 70 mL batch micro scale reactor. The optimum conditions obtained at 430 ◦ C, 90 min, 10 bar, producing the highest conversion of gasoline-like hydrocarbons (∼24.38%), as well as kerosene, light gas oil, gas oil, residues, hydrocarbon gases, and small amounts of solids (∼11.98%, 24.35%, 5.70%, 13.86%, 19.07%, and 0.65%), respectively. The drawbacks of catalytic cracking investigations in micro scale reactors is the small quantities of feed used, producing liquid hydrocarbons only for GC–MS analysis, thus making it not possible to collect enough quantities of OLP necessary to carry out complete physicochemical characterization [29–33,35,43–48]. Pütün et al. [18,20] investigate the pyrolysis of soybean cake (slow) and olive oil residue (fast) using a 316 stainless steel fixed bed laboratory scale reactor (ID = 70 mm, L = 104 mm, VR = 400 mL), heated by an electric furnace, connected to a water-cooled condenser, coupled to a set of traps (collectors) and a gas flow meter. Pütün et al. [18] studied the slow pyrolysis of soybean cake under static, N2 , and steam atmosphere to investigate the Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx effect of temperature (400, 500, 550, and 700 ◦ C) and particle size. The optimum conditions showed a OLP yield of 33.78% (wt.) at 550 ◦ C, for 200 cm3 min−1 N2 , heating rate of 7 ◦ C min−1 , and 0.850 < Dp < 1.250 mm particle size, and a OLP yield of 42.79% (wt.) for steam velocity of 1.3 cm/s. The OLP presents aliphatic, aromatics, and polar substances. Pütün et al. [20] investigated the fast pyrolysis of olive oil residue at 400, 500, 550, and 700 ◦ C, at various heating rates, gas flow velocities, particle sizes, and reaction times on the product yields and OLP composition. The results showed a highest OLP yield of 46.72% (wt.), obtained for 0.85 < Dp < 0.45 mm particle size, heating rate of 500 ◦ C min−1 , gas flow rate of 400 cm3 min−1 , at 500 ◦ C, with OLP products similar to petroleum fuels. Şensöz and Angın [19] investigate the pyrolysis of Safflower seed presscake, at 400 and 600 ◦ C, heating rates of 10, 30 and 50 ◦ C min−1 , and variable N2 flow rate, using a stainless steel fixed bed laboratory scale reactor (ID = 70 mm, L = 104 mm, VR = 400 mL), heated by an electric furnace (600 W), equipped with a 180 mm stainless steel pipe (ID = 45 mm) for the exit of gaseous products, and collecting bottles cooled by a salt–ice–water mixture. The yield of coke, gas and OLP ranged between 25 and 34% (wt.), 19 and 25% (wt.), and 28 and 36% (wt.), respectively. The highest OLP yield obtained at 500 ◦ C, 50 ◦ C/min, and a N2 flow rate of 100 cm3 min−1 . Buzetzki et al. [27] investigated the catalytic cracking of filtered frying oils with acid values between 0.70 and 7.0 mg KOH/g, over 10% (wt.) zeolite (NaY, Clinoptilolite-CL, and HZSM-5), at temperatures between 350 and 440 ◦ C, heating rate of 10 ◦ C min−1 , and 1.0 atm, using a stainless steel laboratory scale tubular neck reactor (ID = 70 mm, L = 104 mm, VR = 400 mL), heated by a rose-shaped GLP burner, with mechanical stirring in batch mode, connected to a series of 03 borosilicate-glass downward condensers, cooled with cold water, coupled to a round bottom two neck borosilicate glass flask collector with outer joints, 24/40 sides, one connected to the end of last condenser and the other used to releases the non-condensable gaseous to the atmosphere. The reactor chamber consists of two parts, a lower one, where the frying oil contacts to the catalyst, and an upper part, where the catalysts form a fixed bed, reacting with the up flowing gaseous products. The temperatures measured by thermocouples inserted in the lower mixing chamber and above the catalyst fixed bed at the top of the reactor. The yield of untreated condensates (OLP + aqueous phase), coke, and non-condensable gases ranged from 85 to 93% (wt.), 3 to 7% (wt.), and 4 to 9% (wt.), respectively. The acid values of OLP after separation of aqueous phase and evaporation of short (C1 , and C2 ) carboxylic acids, ranged between 74 and 116 mg KOH/g, the densities between 0.867 and 0.882 g/cm3 , and kinematic viscosities between 4.451 and 9.202 mm2 s−1 , showing that conversion of acyl esters into hydrocarbons was incomplete. Buzetzki et al. [49] investigated the influence of zeolites (NaY, HY, NH4 Y, Na-ZMS5, H-ZMS5) content on the yields of reaction products (OLP, coke, and gas) and physicochemical properties (viscosity, density, and acid value) of rapeseed oil cracking. The experiments carried out at temperatures between 350 and 440 ◦ C, heating rate of 10 ◦ C/min, and 1.0 atm, using the same apparatus described elsewhere [27], showed that the yield of untreated condensates (OLP + aqueous phase), coke, and non-condensable gaseous products ranged from 75 to 90% (wt.), 4 to 7% (wt.), and 6 to 18% (wt.), respectively. The acid values, densities, and kinematic viscosities of OLP ranged from 98 to 119 mg KOH/g, 0.874 to 0.876 g/cm3 , and 7.259 to 7.598 mm2 s−1 , respectively, showing that conversion of acyl esters into hydrocarbons was not complete, because of high concentration of carboxylic acids in OLP. Dandik and Aksoy [21] investigated the pyrolysis and catalytic cracking of residual sunflower oil over Na2 CO3 , silica-alumina and HZSM-5, using a stainless steel laboratory scale reactor (ID = 45 mm, L = 210 mm, VR = 334 mL), coupled to a fractionation column (ID = 45 mm, L = 540 mm), packed with ceramic rings (ID = 7 mm), obtaining gaseous (CO, CO2 , H2 , CH4 ) and liquid 3 products (hydrocarbons, carboxylic acids, water), coke, and residual oil. The results showed a maximum OLP yield of 32.80% (wt.), using Na2 CO3 at 420 ◦ C, as well as hydrocarbons in the temperature boiling range of gasoline. Dandik and Aksoy [28] studied the catalytic cracking of residual sunflower oil, coupled to a fractionation packed column packed with 03 different heights (ID = 45 mm, L1 = 180 mm, L2 = 360 mm, and L3 = 540 mm), at 400 and 420 ◦ C, with 1, 5, 10, and 20% (wt.) Na2 CO3 , obtaining conversions between 43 and 83% (wt.). The OLP products distribution strongly depends on temperature and Na2 CO3 content. By increasing the catalyst content and temperature, OLP and gas yields increases, while the yield of aqueous phase, acid phase and coke–residual oil decreases. By increasing the column length, the amount of gas and coke–residual oil increases and the amount of liquid hydrocarbon and acid phase decreases. The highest C5 C11 yield of 36.4% (wt.), obtained by using 10% Na2 CO3 , a packed column of 180 mm at 420 ◦ C. The gas products included mostly C1 C3 hydrocarbons. W. H. Chang and C. T. Tye [34], investigated the catalytic cracking of used palm oil using composite zeolites (H-ZSM5, Cu-HZSM5, Zn-HZSM5, Mg-HZSM5), at 350 ◦ C and 1.0 atm, using a batch laboratory scale reactor (Parr 4570), with mechanical stirring, and a cooling coil system. The obtained OLP, gas and residue yields ranged between 75.94 and 84.70% (wt.), 5.63 and 15.80% (wt.), and 8.20 and 10.81% (wt.), respectively. The best result obtained with Mg-HZSM5, producing an OLP, gas and residue yields of 84.70% (wt.), 5.63% (wt.), and 9.64% (wt.), respectively, as well as gasoline and diesel fractions of 9% (vol.) and 72.5% (vol.), respectively. Liew et al. [39], investigated the catalytic cracking of waste chicken over ZSM-5, at 400 ◦ C, using a laboratory scale distillation apparatus under N2 flow, obtaining a bio-oil containing hydrocarbons (C7 C24 ), carboxylic acids, alcohols, ketones, esters, aromatics, anhydrides, ether, and aldehydes. Hua Tian et al. [40] investigate the catalytic cracking of oils (palm and soybean) and chicken fat, at 1.0 atm, using a laboratory scale two-stage riser fluid catalytic cracking, the 1st stage riser at 500 ◦ C, catalysts-to-feed ratio = 6, and 1.4 s residence time, and the 2nd stage riser at 520 ◦ C, catalysts-to-feed ratio = 8, and 1.7 s residence time, over USY and ZSM-5 zeolites. The results showed a LPG yield of 34.34% (wt.), a gasoline yield of 32.75% (wt.), a diesel yield of 11.40% (wt.), a gas yield of 4.48% (wt.), a heavy oil yield of 2.95% (wt.), and a coke yield of 2.31% (wt.). The diesel-like fraction showed an acid value of 190 mg KOH/ml and the kinematic 3.91 mm2 s−1 . Takwa Kraiem et al. [23] investigated the pyrolysis of waste fish fats at 500 ◦ C, 5 ◦ C min−1 , and N2 flow rate of 0.3 cm3 min−1 , using a stainless steel fixed bed bench scale reactor (ID = 150 mm, L = 300 mm, VR = 5300 mL), connected to a cryostat cooling system, and a funnel to separate the bio-oil from aqueous phase. The (OLP + aqueous) yield was 54.60% (wt.), formed by 17.10% (wt.) aqueous phase and 37.50% (wt.) bio-oil with an acid value of 103.14 mg KOH/g, pH of 3.12, and kinematic viscosity (40 ◦ C) of 6.980 mm2 s−1 , showing that conversion of fish fats into hydrocarbons was incomplete. Santos et al. [26] investigated the pyrolysis of industrial fatty waste (soybean soap stock, beef tallow and poultry industry waste) and Lima et al. [50] the pyrolysis and catalytic cracking of soybean, palm, and castor oils over HZSM-5, using a stainless steel bench scale batch reactor of 5.0 L, heated by electrical resistances, connected to a stainless steel water-cooled condenser, coupled to a collector, as well as a deoxygenating tubular borosilicate-glass fixed bed reactor, placed between the reactor and the water-cooling system. The pyrolysis of industrial fatty wastes (soybean soap stock, beef tallow, and poultry industry waste), produced OLP containing hydrocarbons and oxygenates, and the diesel-like hydrocarbons fractions presents alkenes, alkanes olefins, carboxylic acids and esters. The yields of OLP ranged from 8.0 to 32.0% (wt.), being the maximum obtained with beef tallow. The results showed acid values, densities, and kinematic viscosities of OLP ranging from 3.02 (soybean soap stock), 87.07 (beef Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 4 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx tallow), to 114.43 (chicken fat) mg KOH/g, 0.840 to 0.882 g/cm3 , and 3.02 to 4.93 mm2 s−1 , respectively, showing that conversion of acyl esters into hydrocarbons fuels was incomplete, except for the pyrolysis of soybean soap stock [26]. The pyrolysis of soybean and castor oils occurred at temperatures between 350 and 400 ◦ C, while that of palm oil between 330 and 380 ◦ C. The gas reaction products passed through a 2.0 cm bed of HZMS-5, placed inside a fritted bottom borosilicate-glass tube, heated by an external electric resistance to maintain the gaseous reaction products entering the fixed bed reactor at 400 ◦ C. GG-MS analysis identified the presence of hydrocarbons and oxygenates, including alkanes, alkenes, dienes, and carboxylic acids. The acid values, densities, and kinematic viscosities of OLP ranged from 116.20 to 207.50 mg KOH/g, 0.818 to 0.844 g/cm3 , and 2.70 to 3.70 mm2 s−1 , respectively, showing that conversion of acyl esters into hydrocarbons was incomplete [50]. Weber et al. [41] and Bojanowski et al. [42] investigated the thermal and catalytic conversion of animal meal and MBM at 400 ◦ C, using a bench scale Pyrex-glass reactor (L = 140 cm, ID = 7 cm, VReactor ≈ 5400 mL), mounted inside a heat controller furnace of 10.1 kW, connected to a water-cooled condenser, coupled to a separating funnel, and to a gas trap. The results for the thermal conversion of animal meal showed an OLP, solid (coke + salt), water, and gas yield of 29.2%, 46.8%, 14% (wt.), and 10% (wt.), with a density of 0.94 g/ml and kinematic viscosity of 45.2 mm2 /s, indicates the presence of carboxylic acids and non-reacted esters of triglycerides in high concentrations, and an incomplete conversion of animal meal. The results for thermal conversion of MBM show an OLP yield of 16.3% (wt.), with a density of 0.90 g/ml and kinematic viscosity of 29.8 mm2 /s, indicating the presence of carboxylic acids and non-reacted esters of triglycerides in high concentrations, and an incomplete conversion of MBM. NMR analysis of OLP from MBM identified the presence of 97% (area.) aliphatic and 3% (area.) aromatic compounds, although FT-IR analysis of MBM confirms the presence of carboxylic acids and esters of triglycerides. The catalytic conversion of animal fat and MBM carried out in a vertical bench-scale quartz glass reactor (ID = 29 mm, L = 1100 mm, VReactor ≈ 726.5 mL), mounted inside a block controller furnace of 1.8 kW, connected above to a cracking/evaporation vessel of 1.0 L [37,41,42]. The quartz glass reactor with a 30 cm catalyst bed connected to a water-cooled glass condenser, attached to a separating funnel, and to a gas trap. Granular (2–5 mm) Y-type zeolite (Wessalith) and H-type zeolite (Pentasil) investigated at 400 [37,42], and H-type zeolite (Pentasil) at 550 ◦ C [37], and 8.7% (wt.) Na2CO3 with 5% (wt.) H2 O at 410–450 ◦ C [41]. The results for catalytic conversion of animal fat at 400 ◦ C showed an OLP, coke, and gas yield of 72.90%, 5.80%, 16.20% (wt.), respectively, with density of 0.80 g/ml and kinematic viscosity of 2.90 mm2 /s for Y-type zeolite, and an OLP, coke, and gas yield of 56.74%, 0.87%, and 35.22% (wt.), respectively, for H-type zeolite at 400 ◦ C. RMN analysis of OLP obtained from animal fat identified similar composition to that of MBM. The results for catalytic conversion of animal fat at 550 ◦ C showed an OLP yield of 31.5% (wt.), with a density of 0.84 g/ml and kinematic viscosity of 0.74 mm2 /s. The results for the catalytic cracking of animal meal with 8.7% (wt.) Na2CO3, 5% (wt.) H2 O between 410 and 450 ◦ C, showed an OLP, solid (coke + salts), and water yields of 21.61%, 10.42% (wt.), and 11.24% (wt.), respectively, with an acid value of 13.3 mg KOH/g and kinematic viscosity of 14.5 mm2 /s [41]. The distillation fractions of bio-crude from animal meal produces 65.8% (wt.) bio-diesel and 13.3% (wt.) bio-gasoline. NMR and GC–MS analysis identified the presence of toluene, isomeric xylenes, mesi-tylene and isomeric trimethylbenzenes as major compounds. FT-IR analysis identified the presence of carboxylic acids and acyl esters. Meier et al. [22,24,25,51], investigated the fast pyrolysis of waste fish oil [22], waste cooking oil [24,25], and soybean oil [51] in a continuous pyrolysis semi-pilot reactor (ID = 70 mm, L = 2310, VR = 8890 mL), constructed of 12 sections with electrical resistances of 1 kW and temperature control. A feed tank of 20 L, a metering pump with positive displacement connected to a serpentine type pre-heater (30 kW) (L = 600 cm, ID = 2 cm), provided a mass flow rate up to 3.0 kg/h. A cooling system consisting of 02 condensers, a stainless steel shell and tube heat transfer (L = 15.2 cm, ID = 2.82 cm), and a glass condenser (L = 97 cm, ID = 8.0 cm), coupled to a cooling tower of 4.5 m3 /h. Both condensers connected to a gasliquid flash drum of 1.85 L and a collector of 40 L. Fast pyrolysis of waste fish oil carried out at 525 ◦ C and 3.0 kg/h [22]. The yield of bio-oil ranged from 72 to 73% (wt.). The bio-oil distillation fractions were physicochemical characterized and compared to ANP No. 65 [53], showing physicochemical properties and compositions similar to petroleum fuels. Fast pyrolysis of waste frying oil carried out at 475, 525 and 575 ◦ C, 50% (wt.) water, and 3.0 kg/h [24]. The yields of OLP, coke, and gas varied between 56 and 77% (wt.), 0 to 17% (wt.), and 20 to 44% (wt.). The GC identified the presence of hydrocarbons fraction between C4 C10 and C9 C16 after collecting OLP from flash drums 2 and 1, respectively. Fast pyrolysis of waste frying oil carried out at 475, 500, and 525 ◦ C, and feed flow rates between 0.78–3.65 kg/h [25]. The yields of OLP varied between 49.5 and 75.2% (wt.), with acid values between 113.80 and 175.90 [mg KOH/g], and GC identified the presence of hydrocarbon and carboxylic acids in high concentrations. The maximum light bio-oil yield of 78.4% (wt.) obtained at 525 ◦ C and 0.78 kg/h, while the maximum heavy bio-oil yield of 52.4% (wt.) at 475 ◦ C and 2.86 kg/h. Fast pyrolysis of soybean oil carried out at 450, 525 and 600 ◦ C, 0, 5, and 10% (wt.) water, and 3 kg/h [51]. The yields of OLP and gas varied between 34.2 and 91.7% (wt.), and 8.3 and 65.8% (wt.) with highest OLP yield (91.7%) and lowest gas yield (8.3%) obtained at 450 ◦ C with 10% (wt.) water. The GC analysis identified the presence of hydrocarbon fractions between C4 C9 and C10 C28 . The distillation fractions of OLP obtained at 450 ◦ C with 10% (wt.) water, contained 10.74% gasoline, 70.77% green diesel, and 18.49% (wt.) heavy bio-oil. Weber et al. [36,38], investigated the catalytic cracking of a fatty acid mixture (60:40 oleic and stearic acids) and animal fat [36], and the chemical characterization of hydrocarbons from catalytic cracking of animal fat [38], using a continuous moving bed reactor in pilot scale. The extent of decarboxylation increases with increasing contact time and/or the frequency of contact between the coexisting gas-solid phases. The modular reactor with dimensions (L = 6.0 m, B = 2.4 m, H = 3.1 m) and 0.75 m3 [36,54], operates at feed mass flow rate up to 50.0 kg/h. It has an upper drive head for mixing and recirculation, 06 heating systems, insulated with cement bricks. Auxiliary systems includes a feed systems (feed screw, discharge screw, rotary and collecting), feed tank with inlet shaft, condensing system, and a separation unit. A control unit displays the temperature, pressure, weighing, and mass flow rates. The catalytic cracking of fatty acids mixture (60:40 oleic and stearic acids) was carried out at temperatures between 410 and 450 ◦ C and feed mass flow rate of 10 kg/h, using Na2 CO3 as catalyst and 5% (wt.) H2 O [36], while the feed mass flow rate for animal fat varied between 10 and 40 kg/h. The bio-oil yield of fatty acids conversion varied from 64 to 74% (wt.) with an acid value within the range of 0.64 and 0.80 mg KOH/g, while the bio-oil yield of animal fat conversion ranged from 60 to 70% (wt.) with an acid value between 0.5 and 1.80 mg KOH/g, and mean kinematic viscosity of 1.78 mm2 s−1 . The yield of gaseous products ranged from 25 to 30% (wt.), and that of coke between 4 and 6% (wt.). In addition, the distillation of bio-oil yields 66% (wt.) bio-diesel and 21% (wt.) bio-gasoline. The thermochemical conversion of animal fat carried out at 430 ◦ C, with 800 kg Na2 CO3 /40 kg triglyceride per hour and 5% (wt.) H2 O, to produced salts of carboxylic acids [38]. The chemical analysis of bio-oil identified a homologous series of straight chain alkanes and alkenes. Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx In addition, fractionated distillation of the bio-oil yields diesel-like fractions matching nearly all parameters of European Diesel Fuel standards. Almeida et al. [17] and Mota et al. [52] investigated the production of liquid hydrocarbons by catalytic cracking of palm oil and residual fat (FOG) using a pilot scale stirred tank slurry reactor of 143 L, operating in batch mode. The production of liquid hydrocarbons by catalytic cracking of palm oil carried out at 450 ◦ C, 1.0 atm, with 20% (wt.) Na2 CO3. The results showed an OLP yield of 65.86% (wt.), 30.24% (wt.) non-condensable gases, 2.5% (wt.) water, and 1.4% (wt.) coke, with an acid value of 1.02 mg KOH/g, density of 0.790 g/cm3 , and kinematic viscosity of 2.02 mm2 s−1 . The yield of light diesel obtained by distillation average 24.9% (wt.) with an acid value of 1.68 mg KOH/g, density of 0.790 g/cm3 , and kinematic viscosity of 1.48 mm2 s−1 , being composed by 91.38% of hydrocarbons (31.27% normal paraffins 54.44% olefins and 5.67% of naphthenics), and 8.62% of oxygenate compounds [52]. The catalytic cracking of residual fat (FOG) from grease traps carried out at 450 ◦ C, 1 atm, with 5, 10, and 15% (wt.) activated Red Mud as catalyst. The results showed an OLP yield ranging from 62.34 to 75.92% (wt.) with acid values between 84.65 and 109.55 mg KOH/g and kinematic viscosity between 10.96 and 14.08 mm2 s−1 . Distillation of OLP obtained with 5% (wt.) activated Red Mud yield 6.39% (wt.) light diesel-like hydrocarbons fraction with an acid value of 126.24 mg KOH/g and 41.33% (wt.) heavy diesel-like hydrocarbons fraction with an acid value of 94.18 mg KOH/g. Despite several studies on the pyrolysis [18–26,50,51,55] and catalytic cracking [17,27–49,52,56] of crude and refined vegetable oils, and residual fats and oils, detailed information on the physicochemical properties of OLP and distillation fractions are scarce [17,22,25,36–38,41,42,49,50,52]. In addition, only a few studies on pyrolysis and catalytic cracking of fats and oils have been carried out in continuous and/or batch mode in pilot scale [17,36,38,52], and until the moment no study has investigated the performance of catalytic cracking in different production scales. In this context, this study aims to evaluate the effect of catalytic cracking of FOG from grease traps in different scales (bench, laboratory, and pilot) on the yield of reaction products and physicochemical properties of OLP and provide a data basis to perform engineering process analysis (process design, scale-up, and economic evaluation). 2. Materials and methods 2.1. Materials The residual fat, oils, and grease material (FOG) used as renewable raw material, was collected at the grease traps of UFPA Students Restaurant. Solvay Chemicals International SA (Brussels, Belgium) supplied Sodium Carbonate (commercial soda ASH Light D50) with 98.0% (wt.) purity. 2.2. Pre-treatment of residual fat, oils, and grease material (FOG) A previous study has described in details the pre-treatment of residual fat material, a complex mixture consisting of residual fat, oils, and grease (FOG), aqueous phase and suspended solids [17]. 2.3. Physicochemical analysis of residual fat, oils, and grease material (FOG) The dehydrated residual fat material (FOG) has been physicochemical characterized for acid value by AOCS Cd 3d-63, saponification value by AOCS Cd 3-25, free fatty acids by ASTM D5555, density by the pycnometer method ASTM D1480, and kinematic viscosity according to ASTM D445/D446 methods using a 5 Cannon-Fenske viscometer (Schot Geräte, Model: 520 23), with a capillary tube No. 200. 2.4. Catalyst characterization The elemental chemical analysis of Na2 CO3 by XRF has been investigate to determine its purity grade. In a previous study the mineralogical and qualitative chemical analysis of Na2 CO3 has been investigate by XRD, TGA/DTG, and FT-IR [52]. 2.5. Characterization of OLP and distillation fractions 2.5.1. Physicochemical analysis of OLP and distillation fractions The OLP and distillation fractions (kerosene, light diesel, and heavy diesel-like hydrocarbons) obtained by fractional distillation of OLP produced in pilot scale with 10% (wt.) Na2 CO3 have been physicochemical characterized according to the analysis described in Section 2.3. In addition, OLP and distillation fractions were analyzed for corrosiveness to copper (ASTM D130-12) using a copper strip test analyzer (PETROTEST DP, Model: E 25-0600), flash point (ASTM D93) using a Pensky-Martens close cup analyzer (TANAKA, Model: APM-7/FC-7), and refractive index (AOCS Cc 7-25) [17]. 2.5.2. FT-IR analysis of residual fat material (FOG), OLP and distillation fractions A previous study described the FT-IR analysis of residual fat material (FOG) in details [17]. FT-IR analysis of OLP and distillation fractions (kerosene, light diesel, and heavy diesel-like hydrocarbons) obtained by fractional distillation of OLP, produced in pilot scale with 10% (wt.) Na2 CO3 , performed according to the procedures described elsewhere [17,52]. 2.5.3. NMR analysis of OLP and kerosene-like hydrocarbons fraction This study applied the NMR spectroscopy technique of 1 H and 13 C not only for the chemical elucidation but also to better characterize the qualitative chemical composition of OLP and kerosene-like hydrocarbons fraction obtained by fractional distillation of OLP, produced in pilot scale with 10% (wt.) Na2 CO3 . The NMR spectra of 1 H and 13 C were obtained using a spectrometer (Varian, Model UNITY 300), with a resonance frequency of 300 MHz. The solvent was deuterated chloroform (CDCl3) and as a reference standard substance [(CH3)4Si] was used, being the acquisition conditions of 1 H NMR spectra obtained with a pulse at 30 and 32 transients, while that of 13 C NMR spectra obtained with a pulse at 30 and 3940 transients. 2.5.4. GC–MS analysis of OLP and kerosene-like hydrocarbons fraction Prior to the chemical analysis by GC–MS, the fatty acids present in OLP and kerosene-like hydrocarbons fraction were submitted to derivatization. Initially, an aliquot of 20.0 ␮L OLP transferred to a vial. Afterwards, it was added 100 ␮L of N-methyl (trimethylsilyl) trifluoroacetamide (MSTFA). The mixture (OLP + MSTFA) was homogenized and heated at 60 ◦ C for 30 min using an orbital shaker with temperature control. Then, the homogeneous liquid phase diluted in 880 ␮L of CH2Cl2 (solvent) and injected into the GC–MS apparatus [41]. The GC–MS analysis of OLP and kerosenelike hydrocarbons fraction performed as described in a previous study [17]. 2.6. Experimental apparatus and procedures 2.6.1. Laboratory scale unit Fig. 1 describes the schematic diagram of laboratory scale unit. It consists of a cylindrical borosilicate-glass reactor of 150 mm height Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 6 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Fig. 1. Schematic diagram of laboratory scale borosilicate-glass catalytic cracking reactor of 143 mL. and 35 mm diameter with a 34/35 standard taper outer joint connected to a 300 mm jacket length borosilicate glass condenser unit with 24/40 standard taper outer and inner joints. A cylindrical heating system constructed of ceramic resistances of 800 W with a digital temperature control (Therma, Model: TH90DP202-000) and a thermocouple of type K (Ecil, Model: QK-2), with a precision of ±1.0 K jacket the borosilicate-glass reactor. The upper part of the borosilicate-glass adapter with a 34/35 standard taper inner joint connected to the reactor and a 24/40 standard taper outer joint to the condenser unit, consisting of a small size glass tube connected to a N2 flow stream, using a capillary silicone tube from nitrogen gas reservoir. A heating plate with a magnetic stirrer (Ika, C-MAG, Model: HS7), provided agitation with aid a magnetic bar inside the reactor. A glass sampling unit of 250 mL, a cylindrical nitrogen reservoir with a two stage pressure relieve valve (Cemper, Model: CS 54), a gas flow meter (Omel, Model: 189-162), calibrated with air at 1 atmosphere and 294 K, within the range 0–200 mL/min, and a gas exhaust system release the non-condensable gases into the atmosphere. Initially, approximately 48 mL of sample and 10% (wt.) of Na2 CO3 is weighed using a semi-analytical balance (QUIMIS, Model: Q-500L210C). Then, the solid paste dehydrated residual fat was liquefied using a heating plate with a magnetic stirrer (IKA, CMAG, Model: HS7) at 150 ◦ C. Afterwards, the dehydrated residual fat material, the catalyst, and the magnetic bar introduced inside the reactor. Then, the reactor connected to the condenser, cooled with water at 10 ◦ C, inserted inside the cylindrical furnace, and connected to a nitrogen cylinder. The operating temperature was set to 450 ◦ C with a heating rate of 10 ◦ C/min, and a nitrogen flow to 40 mL/min. The reaction time was the time necessary to reach the reactor set point temperature and the cracking temperature defined as the temperature the gaseous products were formed. The collected liquid product placed in a separation funnel to remove the aqueous phase, being the OLP washed (liquid–liquid extraction) three times with distilled water in the proportion 2:1 at 70 ◦ C. The samples were stored in vials for subsequent physicochemical analysis. 2.6.2. Bench scale unit Fig. 2 shows the schematic diagram of bench scale catalytic cracking unit mounted on a mobile metal structure (Laboratory of Catalysis Preparation and Catalytic Cracking-IME). The unit consists of a stainless steel (AISI 304) tubular reactor (ID = 8.5 cm, L = 35.5 cm, VOperational ≈ 1.5 L), inserted inside a thermal blanket of 3.5 kW, isolated by glass wool, a mechanical agitation system, coupled to the reactor. The agitation system constructed by a turbine type impeller with 04 blades at 45◦ and 0.06 kW power, internal impeller diameter of 8.0 cm and maximum angular velocity of 800 rpm. The reactor (PMax = 10 bar, TMax = 550 ◦ C), coupled to a stainless steel double pipe condenser of 1/2′′ nominal diameter with a heat transfer area of 500 cm2 , designed to operate on the shell side (cooling water) (PMax = 10 bar, TMax = 550 ◦ C), and on the tube side (gaseous reaction products) (PMax = 10 bar, TMax = 550 ◦ C). The cooling system, a thermostatic bath with digital temperature control and recirculation, provided cool water at 10 ◦ C. The condensates (OLP + aqueous phase) were collected inside a stainless steel collection vessel of 2 L (PMax = 1.0 bar, TMax = 100 ◦ C), being the operation temperature that of environment. The non-condensable Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS 7 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Fig. 2. Schematic diagram of bench scale stainless steel catalytic cracking reactor of 1.5 L. gaseous products and the inert gas (N2 ) flow into an exhaust valve placed between the shell and tube condenser and the collection vessel, being released to the atmosphere. The feeding of inert carrier gas (N2 ) enters the reactor by a gas inlet port. In addition, the bench scale catalytic cracking unit still contains a stainless steel cylindrical fixed bed reactor of 2 L, inserted inside a thermal blanket of 1.5 kW, isolated by glass wool (PMax = 10 bar, TMax = 550 ◦ C). A digital control unit, containing a LCP, controls the reactor temperature, the frequency inverter (mechanical impeller rpm), the heating rate, and the gas carrier (N2 ) flow rate. The digital control unit displays also the condenser inlet and outlet temperatures. The bench scale catalytic cracking unit also contains a decarboxylation tubular stainless steel (AISI 304) fixed bed reactor (R-02) with 30 cm height and 1.5 cm internal diameter (VOperational ≈ 0.05 L), heated by an electric blanket of 1.5 kW (PMax = 10 bar, TMax = 550 ◦ C). Initially, approximately 740 mL of sample and 72.25 g of Na2 CO3 were weighed in a semi-analytical balance (QUIMIS, Model: Q500L210C). Then, the solid paste dehydrated residual fat was liquefied using a heating plate with a magnetic stirrer (IKA, CMAG, Model: HS7) at 150 ◦ C. Afterwards, the dehydrated residual fat material and the catalyst introduced inside the reactor, and the desired reactor temperature (450 ◦ C), heating rate (10 ◦ C/min), mechanical impeller rpm (600 rpm), and carrier gas (N2 ) flow rate of 0.04 NL/min were set. Then, the condensed gaseous reaction products, collected inside the collection vessel, weighed. Afterwards, the collected condensates (OLP + aqueous phase) were washed (liquid–liquid extraction) three times with distilled water in the proportion 2:1 at 70 ◦ C using a decantation funnel, being the OLP submitted to filtration to remove catalyst particles. Then, the purified OLP and the solid residue (coke) inside the reactor weighed to perform material balances and process yields. The reaction time computed as the time necessary to go from ambient temperature (25 ◦ C), with a heating rate of 10 ◦ C/min, until the reactor reached the set point temperature (450 ◦ C). 2.6.3. Pilot scale unit The apparatus described in details elsewhere [52], operates in batch mode at 450 ◦ C and 1.0 atmosphere, using residual fat material (scum) from grease traps as renewable raw material and 10% (wt.) Na2 CO3 as catalyst. Almeida et al. [17], described in details the experimental apparatus and procedures of fractional distillation of OLP to obtain diesel-like fractions. 2.7. Material balances of catalytic cracking process The steady state mass balance within the stirred tank sludge bed reactor, operating at atmospheric pressure and batch mode, as well as yield of reaction products was described in details elsewhere [17]. The dehydrated residual fat material used as renewable raw material on the thermal catalytic cracking experiments in laboratory, bench, and pilot scales presented acid values varying between 77.41 and 155.42 mg KOH/g. In this context, the thermochemical conversion of fatty acids, defined in terms of acid values and computed by Eq. (1), has been applied as an auxiliary parameter to analyze the process performance in different production scales. Where IARF is the acid value of dehydrated residual fat material, and IAOLP the acid value of OLP. Conversion of FFA [%] = (IARF − IAOLP ) IARF (1) Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 8 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 1 Physicochemical characterization of dehydrated residual fat, oils, and grease (FOG) used as renewable feed on the catalytic cracking experiments at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). Physicochemical Analysis Laboratory Bench Pilot Hasuntree et. al [14] Kanacki [4] SM-01 Kanacki [4] SM-08 ␳ 60 ◦ C [g/dm3 ] ␯ 60 ◦ C [cSt] ␮ 60 ◦ C [cP] Moisture [%] Insoluble Solids [%] Acid Value [mg KOH/g] Saponification Value [mg KOH/g] Peroxide Value [mEq of O2 /kg] Unsaponifiable Matter [%] Ester Index [mg KOH/g] FFA [%] – – – – – 155.42 – – – – 77.71 – – – – – 98.34 157.15 – – 58.81 49.17 0.980 9.40 9.21 12.00 – 72.41 120.19 – – 47.78 36.20 – – – – – 60.38 – – – – 31.06 – – – 18.06 1.22 – – 0.80 4.83 – 41.80 – – – 55.38 2.51 – – 0.60 0.25 – 14.80 SIM-01: restaurant grease, partially processed. The solids and free water were removed but it had not been cooked. SIM-08: unprocessed restaurant grease. Collected from the tops of three separate barrels. Water mostly at the bottom of the barrel. Ester Index = Saponification Value − Acid Value. Table 2 XRF of Na2 CO3 . Oxides/Chemical Elements Na2 CO3 [wt.%] Al2 O3 SiO2 CaO Fe2 O3 P Na Cl Ag  0.186 0.746 0.098 0.105 0.113 97.83 0.538 0.384 100.000 3. Results and discussions 3.1. Physicochemical analysis of dehydrated fat, oils, and grease (FOG) Table 1 presents the physicochemical analysis of residual fat, oils, and grease (FOG) after pre-treatment described elsewhere [17], used as renewable raw material on the thermal catalytic cracking experiments in laboratory, bench, and pilot scales. A previous study described the physical-chemistry analysis of dehydrated residual fat material used on the catalytic cracking experiment in pilot scale [17]. It can be seen that dehydrated residual fat material used on the catalytic cracking experiment in laboratory scale has an acid value of 155.42 mg KOH/g, showing that residual fat material is composed mainly by carboxylic acids, probably due to hydrolysis reactions as well as microorganism degradation of acyl esters, confirmed by the free fatty acids content of 77.71% (wt.). On the other hand, the residual fat material used on the catalytic cracking experiment in bench scale presented an acid value of 98.34 mg KOH/g and a saponification value of 157.15 mg KOH/g, showing that residual fat material is composed by a mixture of carboxylic acids and residual frying oil/hydrogenated fats of almost equal proportion. These results are in agreement with the free fatty acids content of 49.17% (wt.) and the ester index of 58.81 mg KOH/g. 3.2. Characterization of Na2 CO3 3.2.1. XRF analysis The elemental chemical analysis of Na2 CO3 by XRF has been evaluate to determine the main oxides, the chemical elements present on its crystalline phases, but mainly its purity grade. Table 2 shows the quantitative chemical composition of Na2 CO3 by XRF. It may be observed the presence of oxides (Al2 O3 , SiO2 , CaO, and Fe2 O3 ), Na in high concentration with purity of 97.83% (wt.), as well as P, Cl, and Ag present as traces. Fig. 3. Yields of reaction products (OLP, coke, gas, and water) obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). 3.3. Catalytic cracking of residual fat, oils, and grease (FOG) 3.3.1. Process conditions and material balances of catalytic cracking of fat, oils, and grease (FOG) in different production scales The process conditions, material balances, and yields of reaction products (OLP, coke, gas, and water) obtained by catalytic cracking of residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 as catalyst in different production scales (laboratory, bench, and pilot) are shown in Table 3 and Fig. 3. According to Table 1 and FT-IR analysis of dehydrated residual fat, oils, and grease material (FOG) published elsewhere [17], the residual fat, oils, and grease (FOG) used as raw material on the catalytic cracking experiments present acid values between 72.41 and 155.42 mg KOH/g, conferring to the residual fat material different compositions in terms of carboxylic acids content. As described in Section 2.6 and Table 3, the reactor volume and feedstock mass relations using laboratory, bench, and pilot scales are approximately 1:10:1000 and 1:15:750, respectively. The initial cracking temperatures are almost the same for both laboratory (T = 362 ◦ C) and bench (T = 365 ◦ C) scales experiments but smaller for the pilot scale experiment (T = 306 ◦ C). This may be due to the better thermal insulation system of catalytic cracking pilot unit, which uses special refractory bricks, maintaining the external reactor wall temperature at 50 ◦ C, thus avoiding heat losses to the environment. In addition, the mechanical impeller system of pilot unit has a much higher torque compared to that Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS 9 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 3 Process parameters and overall steady state material balances of catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). Process Parameters Na2 CO3 10% [wt.] Cracking Temperature [ C] Mass of Residual Fat [g] Mass of Na2 CO3 [g] Cracking Time [min] Mechanical Stirrer Speed [rpm] Initial Cracking Temperature [ ◦ C] Mas of Liquid Phase (OLP + H2 O) [g] Mass of Coke [g] Mass of OLP [g] Mass of H2 O [g] Mass of Gas [g] Yield of Coke [wt.%] Yield of OLP [wt.%] Yield of H2 O [wt.%] Yield of Gas [wt.%] ◦ Laboratory Bench Pilot 450 46.65 4.665 73 ≈120 362 34.47 4.57 30.51 3.96 7.61 9.79 65.40 8.49 16.32 450 722.50 72.250 120 600 365 509.94 50.72 454.45 55.49 161.84 7.02 62.90 7.68 22.40 450 34800 3480 129 150 306 25066 3386 23166 1900 6238 9.73 66.57 5.46 18.24 Fig. 4. OLP yields obtained by pyrolysis and catalytic cracking of fats and oils carried out in different production scales (laboratory, bench, and pilot), with reactors volumes ranging between 70 and 143,000 mL, and volume scales between 1 and 2042 [22–25,27,35,48,49]. of bench scale. The experimental data illustrated in Table 3 and Fig. 3 presents OLP yields ranging from 62.90 to 66.57% (wt.), showing no differences between OLP yields in laboratory and pilot scales. The results obtained for OLP yield are according to similar studies reported in the literature [17,21–25,27,28,34–39,42,51]. particularly to those of Weber et al. [36,38], who investigated the production of hydrocarbons by catalytic cracking of a mixture from fatty acids and animal fat using Na2 CO3 as catalyst and Almeida et al. [17], who studied the catalytic cracking of residual fat, oils, and grease (FOG) from grease traps with activated Red Mud as catalyst. Fig. 4 illustrates the OLP yields obtained by pyrolysis and catalytic cracking of fats and oils carried out in different production scales (laboratory, bench, and pilot), with reactors volumes ranging between 70 and 143,000 mL, and volume scales between 1 and 2042 [22–25,27,35,48,49]. One may observes that most OLP values lies between 60 and 80% (wt.), with a slight tendency to decrease with increasing reactor volume. The yield of coke ranged from 7.02 to 9.01% (wt.), showing no differences between the yields of coke obtained in laboratory and pilot scales. The results are higher compared to Charusiri and Vitidsant [35] and Hua Tian et al. [40], lower compared to Şensöz and Angın [19], and according to most studies reported in the literature [17,27,34,37,41,42,49,52]. The high yields of coke are probably due Fig. 5. Density and kinematic viscosity of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). to the high concentration of carboxylic acids present in the residual fat materials, being the carboxylic acids probably polymerized within the surface and porous structure of Na2 CO3 , as carboxylic acids are extremely stable to thermal cracking [55,56]. The yield of gaseous products ranged from 16.32 to 22.40% (wt.), showing small differences between laboratory and pilot scales and its highest value for the bench scale experiment. The results are higher compared to Buzetzki et al. [27], Chang and Tye [34], and Weber et al. [42] for the thermal conversion of animal fat, lower compared to Meier et al. [24] and Tian et al. [40], and according to most studies reported in the literature [17,19,36–38,41,42,49,52]. The yield of water ranged from 5.46 to 8.49% (wt.), showing a decrease with increasing reactor volume (laboratory, bench, and pilot scales), being its highest value observed on the laboratory scale experiments. 3.3.2. Physicochemical characterization of OLP Table 4 shows the physicochemical properties of OLP obtained by catalytic cracking of fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in laboratory, bench, and pilot scales. The experimental results show that increasing the reactor volume (laboratory, bench, and pilot scales) causes a linear monotonic decrease on the OLP density, from 0.835 to 0.820 g/cm3 , Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 10 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 4 Physicochemical characterization of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). Physicochemical Properties Laboratory Bench Pilot OLP OLP OLP ␳ [g/cm3 ] Acid Value [mg KOH/g] (IARF −IAOLP ) × 100 [%] IA 0.835 19.08 0.825 10.45 0.820 14.97 0.82–0.85 – 87.72 89.37 79.35 – Refractive Index [−] ␮ [cSt] Flash Point [◦ C] Saponification Value [mg KOH/g] Ester Index [mg KOH/g] Copper Strip Corrosion (1A) 1.46 4.210 – 43.12 24.04 1 1.45 3.280 – 21.25 10.80 1 1.46 3.485 65 24.22 9.25 1 – 2.0−4.5 – – – 1 RF ANP No. 65 ANP: Brazilian National Petroleum Agency, Resolution No. 65 (Specification of Diesel S10). and saponification values) properties are lower compared to those reported by Almeida et al. [17], showing that Na2 CO3 is highly selective and effective to thermochemical convert low quality fats, oils, and grease (FOG) into hydrocarbons. Fig. 6. Acid and saponification values of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). as illustrated in Fig. 5. The density values are lower compared to similar studies reported in the literature [27,26,37,41,42,49,50], and according to results obtained in pilot scale reported elsewhere [17,52]. The kinematic viscosity and acid and saponification values of OLP exhibits a similar parabolic behavior, showing a minimum at bench scales, as illustrated in Figs. 5 and 6. This may be to the higher mechanical impeller speed (600 rpm). The OLP kinematic viscosity ranged from 4.21 to 3.28 [cSt]. The results are lower compared to Buzetzki et al. [27,49] and Weber et al. [42] for thermal conversion of MBM, but higher compared to Weber et al. [36,37,41,42], and according to Lima et al. [26]. The acid values ranged from 19.08 to 10.45 mg KOH/g, and the saponification values from 43.12 to 21.22 mg KOH/g. The results are higher compared to Weber et al. [36], much lower than most studies similar studies reported in the literature [22–27,37,40–42,49–51], and according to results obtained in pilot scale reported elsewhere [17,52]. Sodium carbonate has not only proven its high performance end selectivity to convert large amounts of fatty acids and acyl esters into hydrocarbons, but also to decrease the acid value of OLP, a fundamental physicochemical parameter conferring the quality of biofuels. In addition, the conversion of fatty acids computed by Eq. (1) and illustrated in Fig. 7 ranged from 89.37 to 79.35%, showing an almost linear monotonic decreases with increasing the reactor volume (laboratory, bench, and pilot scales), as well as small differences on the performance of laboratory and bench scales to thermochemical convert fatty acids into hydrocarbons. The measured OLP physical (density and kinematic viscosity) and physicochemical (acid 3.3.3. Fractional distillation of OLP 3.3.3.1. Mass balances and yields of fractional distillation of OLP. Table 5 shows the mass balances and yields (distillates and raffinate) of fractional distillation of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. It has not been observed the production of gasoline-like hydrocarbons fraction, showing that Na2 CO3 was not selective for hydrocarbon fractions in the temperature boiling point range of gasoline (40 ◦ C < T < 175 ◦ C), which is according to the results of Weber et al. [38]. However, it was observed in all distillation fractions of PLO the presence of hydrocarbon fractions in the temperature boiling point range of kerosene (175 ◦ C < T < 235 ◦ C), light diesel (235 ◦ C < T < 305 ◦ C) and heavy diesel-like hydrocarbons (305 ◦ C < T < 400 ◦ C). The yields of kerosene, light diesel, and heavy diesel-like hydrocarbons were 14.90% (wt.), 32.01% (wt.), and 19.35% (wt.), respectively, while the yield of raffinate was 32.25% (wt.), being the global yield of biofuels equal to 66.26% (wt.). The results are lower compared to similar studies reported in the literature [22,25,34,41,51], higher compared to Dandik and Aksoy [28] and Tian et al. [40], and in accord to data obtained in pilot and laboratory scale reported elsewhere [17,35,52]. Fig. 7. Conversion of fatty acids by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in different production scales (laboratory, bench, and pilot). Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS 11 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 5 Mass balances and yields (distillates and raffinate) of fractional distillation of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. Distillation: Vigreux Column of 03 Stages Mass [g] OLP [g] 643.08 Gas [g] 9.58 Raffinate [g] 207.39 Distillates [g] Yield [wt.%] Boiling Temperature K LD HD K LD HD TI [◦ C] TF [◦ C] 95.82 205.85 124.44 14.90 32.01 19.35 196 396 TI : Initial Boiling Temperature, TF : Final Boiling Temperature, K = Kerosene, LD = Light Diesel, HD = Heavy Diesel. 3.3.3.2. Physicochemical properties of distillation fractions of OLP. Distillation fractions (kerosene-like hydrocarbons fraction: 175–235 ◦ C, light diesel-like hydrocarbons fraction: 235–305 ◦ C, heavy diesel-like hydrocarbons fraction: 305–400 ◦ C) of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale were physicochemical characterized and the results shown in Table 6. Physicochemical analysis presented in Table 6 shows that acid and saponification values, as well as kinematic viscosity and density of kerosene-like, light diesel-like, and heavy diesel-like hydrocarbons fractions decrease with increasing boiling point temperature, showing that OLP still contains carboxylic acids and saponifiable matter (fats) derivate of carboxylic acids, and these compounds have higher densities and kinematic viscosities than hydrocarbons. Thus, an increase on the acid (carboxylic acids) and saponification (fats) values of a distillation fractions, contributes to an increase on its density and viscosity. The results are in accord to similar data obtained in pilot scale reported elsewhere [17,35,52]. The measured values for the density of kerosene, light diesel, and heavy diesel-like hydrocarbons fractions are lower than those within the limits of ANP 37 [57] and 65 [53]. This is probably due to the presence of gasoline heavy compounds (C10 ) in kerosene-like hydrocarbons fraction, as well as kerosene heavy compounds in light and heavy diesel-like hydrocarbons fractions, as confirmed by GC–MS analysis in Table 8. It would be necessary a new rectification step to match the physicochemical properties of Aviation Kerosene specification of ANP No. 37 [57], as well as Diesel S10 specification of ANP No. 65 [53], for density, kinematic viscosity, acid value, and flash point. 3.4. FT-IR analysis of OLP and distillation fractions 3.4.1. FT-IR analysis of OLP Fig. 8 illustrates the FT-IR analysis of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in laboratory and pilot scales. The identification of absorption bands/peaks was done according to previous studies [17,23,50,52]. The spectra of OLP produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales exhibit intense peaks at 2923 cm−1 and 2852 cm−1 , indicating the presence of aliphatic compounds, associated to methylene (CH2 ) and methyl (CH3 ) groups, confirming the presence of hydrocarbons [17,23,50,52], as well as the presence of an intense axial deformation band, characteristic of carbonyl (C O) groups, between 1620 and 1830 cm−1 , being the peaks at 1717 cm−1 probably associated to ketones and/or carboxylic acids [17,23,52]. Although, FT-IR is a qualitative chemical analysis, one observes that the axial deformation band between 1620 and 1830 cm−1 is not only broader, but also the peak intensity at 1713 cm−1 is higher, for OLP produced with 10% (wt.) Na2 CO3 in laboratory scale, probably due to its higher acid value, as described in Table 4. In addition, the spectra of OLP produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales also presents an intense and large axial deformation band between 3200 and 2500 cm−1 , indicating the presence of a hydroxyl (O H) group, characteristics of carboxylic acids [17,23,50,52]. Once again, one might observes that Fig. 8. FT-IR of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in laboratory and pilot scales. the axial deformation band between 3200 and 2500 cm−1 is larger for OLP produced with 10% (wt.) Na2 CO3 in laboratory scale, probably due to its higher acid value, as described in Table 4. The spectra of OLP produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales exhibits between 1456 and 1461 cm−1 , a characteristic asymmetrical angular deformation of methylene (CH2 ) and methyl (CH3 ) groups, indicating the presence of alkanes [17,23,52]. In addition, the spectra of OLP produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales identify in the region between 1378 and 1382 cm−1 , bands of symmetrical angular deformation of C H bonds in methyl group (CH3 ) [17,52]. The peaks at 992 cm−1 , 966 and 909 cm−1 , are characteristic of an angular deformation outside the plane of C H bonds, indicating the presence of alkenes [17,23,52]. The spectra of OLP produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales exhibits bands between 721 and 724 cm−1 , characteristic of an angular deformation outside the plane of C H bonds in methylene (CH2 ) group, indicates the presence of olefins [17,23,52]. The analysis of OLP spectra produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales identify the presence of aliphatic groups (alkenes, Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 12 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 6 Physicochemical characterization of distillation fractions (kerosene-like hydrocarbons fraction: 175–235 ◦ C, light diesel-like hydrocarbons fraction: 235–305 ◦ C, heavy diesellike hydrocarbons fraction: 305–400 ◦ C) of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. Physicochemical Properties Na2 CO3 10 [wt.%] ANP No. 37 Kerosene ␳ [g/cm3 ] I. A [mg KOH/g] I. R [−] ␮ [cSt] I. S [mg KOH/g] Ester Index [mg KOH/g] FP [◦ C] C (1 A) 0.740 5.43 1.427 0.660 9.23 3.90 37 1 0.77–0.83 max 0.015 Note max 8 Note – 38–40 1 Na2 CO3 10 [wt.%] ANP No. 65 Light Diesel Heavy Diesel 0.791 6.79 1.442 1.930 11.90 5.11 38 1 0.812 7.27 1.466 3.510 27.91 20.64 38 1 0.82–0.85 Note Note 2.0–4.5 Note – 38 1 I.A = Acid Value, I.R = Refractive Index, I.S = Saponification Value, FP = Flash Point, C = Copper Corrosiveness. ANP: Brazilian National Petroleum Agency, Resolution No. 37 (Specification of Aviation Kerosene). istic of carbonyl (C O) groups, probably associated to ketones and/or carboxylic acids [17,23,52]. One observes that width of axial deformation band between 1620 and 1830 cm−1 , with peak intensity at 1717 cm−1 , increases with increasing temperature boiling point, probably due to an increase on the acid values of distillation fractions, as described in Table 4. In addition, all the spectra of OLP distillation fractions presents a broad axial deformation band between 3200 and 2500 cm−1 , indicating the presence of a hydroxyl (O H) group, characteristics of carboxylic acids [17,23,50,52]. All the spectra of OLP distillation fractions exhibits a peak at 1461 and 1460 cm−1 , a characteristic asymmetrical angular deformation of methylene (CH2 ) and methyl (CH3 ) groups, indicating the presence of alkanes [17,23,52]. All the spectra of OLP distillation fractions presents peaks at 1380 and 1381 cm−1 , characteristic of symmetrical angular deformation of C H bonds in methyl group (CH3 ) [17,52]. In addition, all the spectra of OLP distillation fractions exhibits peaks at 909 cm−1 , being the presence of peaks at 992 cm−1 and 966, observed only for heavy diesel-like hydrocarbons fraction. These peaks, characteristic of an angular deformation outside the plane of C H bonds, indicate the presence of alkenes [17,23,52]. The peaks at 722 cm−1 and 721, characteristic of an angular deformation outside the plane of C H bonds in methylene (CH2 ) group, indicating the presence of olefins [17,23,52], occurred only for light and heavy diesel-like hydrocarbons fractions. The FT-IR analysis of OLP distillation fractions identify the presence of hydrocarbons, as major chemical compounds, as well as oxygenates, and the absence of aromatic groups. Fig. 9. FT-IR of OLP distillation fractions (kerosene-like, light diesel-like, and heavy diesel-like hydrocarbons fractions) obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. alkanes, etc.), as major chemical compounds, as well as oxygenates (carboxylic acids, ketones, etc.), and the absence of aromatic groups. 3.4.2. FT-IR analysis of distillation fractions Fig. 9 illustrates the FT-IR spectra of OLP distillation fractions (kerosene-like hydrocarbons fraction: 175–235 ◦ C, light diesel-like hydrocarbons fraction: 235–305 ◦ C, heavy diesel-like hydrocarbons fraction: 305–400 ◦ C) obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. The identification of absorption bands/peaks was done according to previous studies [17,23,50,52]. All the spectra of OLP distillation fractions exhibit bands between 2959 and 2923 cm−1 , and 2923 and 2852 cm−1 , with intense peaks at 2958 cm−1 , 2922 cm−1 and 2852 cm−1 , associated to methylene (CH2 ) and methyl (CH3 ) groups, indicating the presence of aliphatic compounds [17,23,50,52]. One also observes the presence of peaks at 1717 cm−1 and 1718 cm−1 , character- 3.5. NMR analysis of OLP and kerosene-like hydrocarbons fraction Figs. 10–13 illustrate the 13 C and 1 H NMR spectra of OLP and kerosene-like hydrocarbons fraction produced by fractional distillation of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. By analyzing the 13 C NMR spectra of OLP and kerosene-like hydrocarbons fraction, one may observe chemical shifts characteristic of CH2 (methylene) and CH3 (methyl) groups of linear and long carbon chain between 14.01–33.81 ppm and 14.00–33.80 ppm, for OLP and kerosene-like hydrocarbons fraction, respectively, as described in Table 7, indicating the presence of aliphatic hydrocarbons. The 13 C NMR spectra also exhibits characteristic signals of olefins hydrocarbons with chemical shifts of carbon double bonds observed at 114.03 and 139.32 ppm for both OLP and kerosene-like hydrocarbons fraction. The 1 H NMR spectrum shows chemical shifts of hydrogen bound to carbon with peaks between 0.87–1.38 ppm and 0.88–1.39 ppm, for OLP and kerosene-like hydrocarbons fraction, respectively, confirming the presence of aliphatic hydrocarbons, as well as chemical shifts of hydrogen and methine ( CH–) group Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 13 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx Table 7 Chemical shifts of 13 C NMR and 1 H NMR characteristics of chemical bonds of compounds presented in OLP and kerosene-like hydrocarbons fraction (175–235 ◦ C) of OLP obtained catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. Kerosene, 10% (wt.) Na2 CO3 OLP, 10% (wt.) Na2 CO3 Chemical Shifts of 13 C NMR Chemical Shifts [ppm] 14.01 ppm 22.66-43.81 ppm 76.76–77.19 ppm 114.03 and 139.32 ppm Chemical Shifts of 1 H NMR Chemical Shifts [ppm] 0.87 ppm for H(F) 1.25 ppm for H(E) 2.12–2.40 ppm for H(D) 4.91–5.00 ppm for H(B) 5.38 ppm for H(A) 5.78–5.83 ppm for H(C) 7.25 ppm Type of Chemical Bond of 13 C R1 CH3 R1 CH2 -R2 (CDCl3 ) Olefins (C C) Chemical Shifts [ppm] 14.00 ppm 22.67-43.80 ppm 76.76–77.18 ppm 114.03 and 139.32 ppm Type of Chemical Bond of 13 C R1 CH3 R1 CH2 -R2 (CDCl3 ) Olefins (C C) Type of Chemical Bond of 1 H R1 CH3 (F) R1 CH2 (E)-R2 Chemical Shifts [ppm] 0.90 ppm for H(F) 1.28 ppm for H(E) 2.04–2.43 ppm for H(D) 4.93–5.02 ppm for H(B) 5.43 ppm for H(A) 5.81–5.86 ppm for H(C) 7.27 ppm Type of Chemical Bond of 1 H R1 CH3 (F) R1 CH2 (E)-R2 (CDCl3 ) (CDCl3 ) Table 8 Class of compounds, summation of peak areas, and retention times of chemical compounds identified by CG-MS of OLP and kerosene-like hydrocarbons fraction obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. OLP, 10% [wt.] Na2 CO3 Class of Compounds: Chemical Compounds Kerosene-like Fraction, 10% [wt.] Na2 CO3 RT [min] Class of Compounds: Chemical Compounds RT [min] Alkenes 1-Decene 1-Undecene 1-Tetradecene 2-Dodecene Tridecene 4.44 5.58 6.68 6.82 7.76 Alkenes 1-Decene 2-Decene 1-Dodecene  (Area.%)= Alkanes 4.44 4.58 5.58 44.99 1-Pentadecene 1-Heptadecene 9-Icosene 3-Heptadecene 8.91 10.12 11.39 12.57 n-Decane Undecane  (Area.%)= Ring containing-Alkenes 4.54 5.67 29.61 1-Nonadecene  (Area.%)= Alkanes n-Decane n-Undecane n-Dodecane n-Tridecane Hexadecane n-Hexadecane Octadecane Heptadecane Docosane  (Area.%)= Ketones 2-Heptadecanone 2-Pentadecanone Nonadecanone  (Area.%)= Carboxylic Acids Dichloroacetic acid Hexadecanoic acid (9Z)-Octadecenoic acid  (Area.%)= Ring containing-Alkenes 1-Butylcyclohexene 1-Heptyl-1-cyclohexene  (Area.%)= Ring containing-Alkanes 1,2-Dibutylcyclopropane 1-Pentyl-2-propylcyclopropane Nonylcyclopentane 1,3-Dicyclohexylpropane  (Area.%)= Fatty Alcohols Cis-9-octadecen-1-ol 1-Heptadecanol  (Area.%)= Dienes 1,19-Icosadiene  (Area.%)= 14.02 39.44 1-Propyl-1-cyclohexene 1-Butylcyclohexene  (Area.%)= Ring containing-Alkanes 1-Hexyl-2-propylcyclopropane 1-Pentyl-2-propylcyclopropane  (Area.%)= Cycloalkanes Cyclohexane Cyclopropane  (Area.%)= Carboxylic Acids Hexanoic acid  (Area.%)= Aromatics Methylbenzene  (Area.%)= 4.12 5.27 7.58 4.54 5.67 6.77 7.85 8.99 10.21 11.47 12.78 14.13 30.91 15.75 19.91 23.46 6.98 4.69 5.72 6.15 4.99 6.68 4.31 5.12 5.38 6.00 1.98 12.69 18.49 22.71 12.14 5.27 11.19 1.50 5.73 5.83 9.65 10.96 4.12 12.41 12.83 1.90 12.49 2.01 Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 14 76.98 76.76 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx 77.19 1.0 0.9 0.8 Intensity 0.7 0.6 0.5 0.4 220 200 180 160 140 120 100 Chemical Shift (ppm) 80 60 22.66 13.87 14.01 0.1 43.81 42.80 33.80 139.32 114.03 0.2 29.68 31.9029.64 29.34 0.3 40 20 0 -20 1.25 Fig. 10. 13 C NMR spectra of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. 1.0 0.9 0.8 0.6 0.87 Intensity 0.7 0.5 16 14 12 10 8 6 Chemical Shift (pp m) 4 2 0.82 0.85 1.30 2.12 1.59 1.38 2.39 2.36 4.91 2.40 5.00 4.97 4.93 0.1 5.78 5.38 5.83 0.2 5.81 5.80 7.25 1.28 0.3 0.86 0.4 0 -2 Fig. 11. 1 H NMR spectra of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. with peaks between 2.12–2.40 ppm and 2.15–2.43 ppm, for OLP and kerosene-like hydrocarbons fraction, respectively. The chemical shifts between 4.91–5.83 ppm and 4.93–5.83 ppm, for OLP and kerosene-like hydrocarbons fraction, respectively, indicate the presence of hydrogen bonded to an unsaturated carbon of olefins in OLP and kerosene-like hydrocarbons fraction. One may also note the absence of peaks between 155 and 185 ppm in the 13 C RMN spectrum, and from 12 to 11 ppm in the 1 H NMR spectrum, both associated to the carbonyl (C O) and hydroxyl (O H) groups, characteristics of carboxylic acids. This confirms the low acid values of OLP and kerosene-like hydrocarbons fraction, as described in Tables 4 and 6. Table 7 Chemical shifts of 13 C NMR and 1 H NMR characteristics of chemical bonds of compounds presented in OLP and kerosenelike hydrocarbons fraction (175–235 ◦ C) of OLP obtained catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 15 77.18 76.76 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx 1.0 0.9 0.8 0.6 14.00 Intensity 0.7 0.5 0.1 220 200 180 160 140 120 100 Chemical Shift (ppm) 80 60 40 20 13.85 14.29 31.91 29.35 0.2 33.80 139.30 0.3 22.67 114.02 0.4 0 -20 0.90 1.28 Fig. 12. 13 C NMR of kerosene-like hydrocarbons fraction of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. 1.0 0.9 0.8 0.88 1.29 0.6 0.5 1.32 Intensity 0.7 0.4 14 13 12 11 10 9 8 5 3 1 0.87 0.92 2.05 2 1.39 2.04 2.28 2.15 4 2.37 4.95 4.93 4.99 5.82 7 6 Chemical Shift (ppm) 2.43 0.1 5.43 5.02 7.27 0.2 5.83 0.3 0 -1 -2 Fig. 13. 1 H NMR of kerosene-like hydrocarbons fraction of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. 3.6. GC–MS analysis of OLP and kerosene-like hydrocarbons fraction The GC–MS analysis of OLP and kerosene-like hydrocarbons fraction obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale are shown in Figs. 14 and 15. Table 8 describes the class of compounds, the summation of peak areas for a class of compounds, and retention times of chemical compounds identified by CG-MS. The GC–MS analysis identified approximately 40 compounds, classified into two major groups (hydrocarbons and oxygenates). The main classes of hydrocarbons detected by GC–MS were alkenes, alkanes, ring-containing alkenes, and ring-containing alkanes, and cycloalkanes, while the main classes of oxygenates were carboxylic acids, ketones, fatty alcohols, and dienes. The OLP presents on its composition 75.97% (area) hydrocarbons and 24.03% (area) oxygenates. The hydrocarbon were composed by 39.44% alkenes (area), 30.91% alkanes (area), 1.50% ring-containing alkanes (area), and 4.12% ringcontaining alkenes (area), while the oxygenates were composed by 12.14% carboxylic acids (area), 6.98% ketones (area), 1.90% fatty alcohols (area), and 2.01% dienes (area). The concentration of carboxylic acids identified by GC–MS in OLP is according to the measured OLP acid value of 14.97 [mg KOH/g]. The results confirm that Na2 CO3 was effective and selective to thermochemical convert fatty acids and lipid base compounds present in fat, oils, and grease (FOG) from grease traps into hydrocarbons, particularly the fatty acids as shown in Table 4, compared to activated Red Mud, as reported elsewhere [17]. GC–MS analysis of OLP identified the presence of carboxylic acids, including Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model ARTICLE IN PRESS JAAP-3716; No. of Pages 18 16 H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx 16 (x10,000,000) 4.5 T IC 13 4.0 17 11 12 3.0 20 4 14 21 2 26 1 5 9 8 2.5 2.0 25 3.5 6 30 1.5 5.0 7.5 10.0 12.5 15.0 34 33 32 28 29 22 2274 2 3 18 19 3 10 0.5 15 7 31 1.0 17.5 20.0 22.5 25.0 27.5 30.0 Fig. 14. GC–MS of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. (x10,000,000) 3 2 T IC 3.5 3.0 2.5 9 2.0 10 1.5 8 6 5.0 12 13 7 1 5 0.5 11 4 1.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 Fig. 15. GC–MS of kerosene-like hydrocarbons fraction of OLP obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale. C15 H32 O (DCA), a product of chlorination of drinking water, produced by the microbiological metabolism of chlorinecontaining compounds such as NaClO, a common disinfectant in food services, with 6.84% (area). Hexadecanoic and (9Z)Octadecenoic acids, the dominating carboxylic acids in soybean and palm oils, as well as hydrogenated fats, were also present in OLP with 3.08 and 2.22% (area), respectively. The kerosene-like hydrocarbons fraction of PLO obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale presents on its composition 94.62% (area) hydrocarbon and 5.38% (area) oxygenates. The hydrocarbon were composed by 44.99% alkenes (area), 29.61% alkanes (area), 7.58% ring-containing alkenes (area), 6.15% ring-containing alkanes (area), 4.31% cycloalkanes (area), and 1.98% (area) aromatics, while the oxygenates were composed by 5.38% carboxylic acids (area). The concentration of carboxylic acids identified by GC–MS in kerosene-like hydrocarbons fraction is consistent to its acid value of 5.43 [mg KOH/g]. The hydrocarbons identified in OLP by GC–MS present carbon chain length ranging from C10 to C22 as follows: alkenes C10 C19 , alkanes C10 C22 , ring-containing alkenes C10 C13 , and ring-containing alkanes C11 C15 , indicating the presence of heavy gasoline compounds with C10 (C5 C10 ), kerosene-like fractions Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx (C11 C12 ), light diesel-like fractions (C13 C17 ), and heavy diesellike fractions (C18 C25 ), according to the definition of fossil fuels fractions [58]. In addition, kerosene-like hydrocarbons fraction presents carbon chain lengths ranging from C9 to C12 with following carbon chain lengths, alkenes C10 C12 , alkanes C10 C11 , ring-containing alkenes C9 C10 , ring-containing alkanes C11 C12 , cycloalkanes C10 C12 , and aromatics with C7 . Once again, one may observe the presence of heavy gasoline compounds C9 C10 in kerosene-like hydrocarbons fraction. This is probably due to the limitation of fractional distillation apparatus (Vigreux Column of 03 Stages). The presence of heavy gasoline compound has probably contribute to a deviation of 3.89%, compared to the lower density limit of ANP 37 [57], as well as a deviation of 2.63%, compared to the lower flash point limit of ANP 37 [57]. 4. Conclusions The dehydrated residual fat material used on the catalytic cracking experiment in laboratory scale contains mainly carboxylic acids, while that used on the experiment in bench scale is a mixture of carboxylic acids and residual frying oil/hydrogenated fats of almost equal proportion. The initial cracking temperatures are almost the same for both laboratory and bench scales experiments but smaller for the pilot scale experiment, probably due to the better thermal insulation of pilot unit. OLP yields ranged from 62.90 to 66.57% (wt.), showing no differences between OLP yields in laboratory and pilot scales. The OLP yields obtained by pyrolysis and catalytic cracking of fats and oils carried out in different production scales (laboratory, bench, and pilot), with reactors volumes ranging between 70 and 143,000 mL, stays between 60 and 80% (wt.), showing a slight tendency to decrease with increasing reactor volume. The experimental results show that increasing the reactor volume (laboratory, bench, and pilot scales) causes a linear monotonic decrease on the OLP density, from 0.835 to 0.820 g/cm3 . The kinematic viscosity and acid and saponification values of OLP exhibits a similar parabolic behavior, showing a minimum at bench scale. This may be to the higher mechanical impeller speed (600 rpm), and confirms the importance of mechanical agitation processes to improve reactor performance. Sodium carbonate has not only proven its high performance end selectivity to convert huge amounts of fatty acids and acyl esters into hydrocarbons, but also to decrease the acid value of OLP, a fundamental physicochemical parameter conferring the quality of biofuels. Distillation fractions of PLO identified the presence of hydrocarbon fractions in the temperature boiling point range of kerosene, light diesel and heavy diesel-like hydrocarbons fractions. The physicochemical properties (acid and saponification values, kinematic viscosity, and density) of distillation fractions decrease with increasing temperature boiling point, showing that OLP still contains carboxylic acids. Densities of kerosene, light diesel, and heavy diesel-like hydrocarbons fractions are lower compared to the lower density limits of ANP 37 [57] and 65 [53], probably due to the presence of gasoline heavy compounds (C10 ) in kerosenelike hydrocarbons fraction, as well as kerosene heavy compounds in light and heavy diesel-like hydrocarbons fractions, as confirmed by GC–MS analysis in Table 8. It would be necessary a new rectification step to match the physicochemical properties of Aviation Kerosene specification of ANP No. 37 [57], as well as Diesel S10 specification of ANP No. 65 [53], for density, kinematic viscosity, acid value, and flash point. The FT-IR analysis of OLP spectra produced with 10% (wt.) Na2 CO3 in laboratory and pilot scales identify the presence of aliphatic groups (alkenes, alkanes, etc.), as major chemical compounds, as well as oxygenates (carboxylic acids, ketones, etc.), and 17 the absence of aromatic groups. The FT-IR analysis of OLP distillation fractions obtained by catalytic cracking of dehydrated residual fat, oils, and grease (FOG) from grease traps at 450 ◦ C and 1.0 atm, with 10% (wt.) Na2 CO3 in pilot scale identify the presence of hydrocarbons, as major chemical compounds, as well as carboxylic acids, in agreement with NMR and GC–MS analysis. The GC–MS analysis of OLP and kerosene-like hydrocarbons fraction obtained in pilot scale with 10% (wt.) Na2 CO3 identified in OLP 76.97% hydrocarbons (39.44% alkenes, 31.91% alkanes, 4.12% ring-containing alkenes, and 1.50% ring-containing alkenes) and 23.03% oxygenates (12.14% carboxylic acids, 6.98% ketones, 1.90% fatty alcohols, and 2.01% dienes). The kerosene-like hydrocarbons fraction is composed by 94.62% (area) hydrocarbon (44.99% alkenes, 29.61% alkanes, 7.58% ring-containing alkenes, 6.15% ringcontaining alkanes, 4.31% cycloalkanes, and 1.98% aromatics) and 5.38% (area) oxygenates (5.38% carboxylic acids), showing that catalytic cracking of scum from grease traps with 10% (wt.) Na2 CO3 is technically feasible. Acknowladgments ELETROBRÁS S/A for the Project financial support. The first author would like to express his gratitude to Prof. Dr. Luiz E. P. Borges (Examiner), for introducing the author into the marvelous field of catalysis and thermal catalytic cracking of vegetable oils, and to Prof. Dr.-Ing Nélio T. Machado (Supervisor), for giving the opportunity to Research as Assistant at the Laboratory of Thermal Separation Processes and Applied Thermodynamics. References [1] T. Aziz, L. Holt, K. Keener, J. Groninger, J. Ducoste, Performance of grease abatement devices for removal of fat oil, and grease, J. Environ. Eng. 137 (1) (2011) 84–92. [2] Uwe Stoll, Hani gupta management strategies for oil and grease residues, Waste Manage. Res. 15 (January) (1997) 23–32. [3] N. Eddy, Restaurants and commercial facilities present specific problems for onsite systems, Small Flow J. 12 (2) (1998) 1–5. [4] M. Canakci, The potential of restaurant waste lipids as biodiesel feedstocks, Bioresour. Technol. 98 (2007) 183–190. [5] N.H. Wong, P.L. Law, S.H. Lai, Field tests on a grease trap effluent filter, Int. J. Environ. Sci. Technol. 4 (3) (2007) 345–350. [6] Zhong-Ming Wang, Jin-Suk Lee, Ji-Yeon Park, Chuang-Zhi Wu, Zhen-Hong Yuan, Optimization of biodiesel production from trap grease via acid catalysis, Korean J. Chem. Eng. 25 (4) (2008) 670–674. [7] J. Hunter Long, Tarek N. Aziz, Francis L. de los Reyes III, Joel J. Ducoste, Anaerobic co-digestion of fat, oil, and grease (FOG): a review of gas production and process limitations, Process Saf. Environ. Prot. 90 (2012) 231–245. [8] Iman A.F. Husain, Ma’an Fahmi Alkhatib, Mohamed Saedi Jammi, Mohamed E.S. Mirghani, Zaki Bin Zainudin, Asif Hoda, Problems, control, and treatment of fat, oil, and grease (FOG): a review, J. Oleo Sci. 63 (No. 8) (2014) 747–752. [9] Robert A. Dangel, David Astraukis, Jean Palmateer, Fatty acid separation from hydrolyser wastewater by ultrafiltration, Environ. Prog. 14 (1) (1995) 65–68. [10] N.G. Wakelin, C.F. Forster, An investigation into microbial removal of fats, oils and greases, Bioresour. Technol. 59 (1) (1997) 37–43. [11] M. Canakci, J. Van Gerpen, Biodiesel from oils and fats with high free fatty acids, Trans. Am. Soc. Agric. Eng. 44 (2001) 1429–1436. [12] Walterio Diaz-Felix, Mark R. Riley, Werner Zimmt, Michael Kazz, Pretreatment of yellow grease for efficient production of fatty acid methyl esters, Biomass Bioenergy 33 (Issue 4) (2009) 558–563. [13] M.J. Montefrio, T. Xinwen, J.P. Obbard, Recovery and pre-treatment of fats, oil and grease from grease interceptors for biodiesel production, Appl. Energ. 87 (2010) 3155–3161. [14] Parichart Hasuntree, Vichuta Toomthong, Sukanda Yoschoch, Usarat Thawornchaisit, The potential of restaurant trap grease as biodiesel feedstock, Songklanakarin J. Sci. Technol. 33 (5) (2011) 525–530. [15] Andres Donoso-Bravo, Maria Fdz-Polanco, Anaerobic co-digestion of sewage sludge and grease trap: assessment of enzyme addition, Process Biochem. 48 (Issues 5–6) (2013) 936–940. [16] J.C. Kabouris, U. Tezel, S.G. Pavlostathis, M. Englemann, J. Dulaney, R.A. Gillette, A.C. Todd, Methane recovery from the anaerobic co-digestion of municipal sludge and FOG, Bioresour. Technol. 100 (15) (2009) 3701–3705. [17] H. da Silva Almeida, O.A. Corrêa, J.G. Eid, H.J. Ribeiro, D.A.R. de Castro, M.S. Pereira, L.M. Pereira, A. de Andrade Mâncio, M.C. Santos, J.A. da Silva Souza, Luiz E.P. Borges, N.M. Mendonça, N.T. Machado, Production of biofuels by Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017 G Model JAAP-3716; No. of Pages 18 18 [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] ARTICLE IN PRESS H. da Silva Almeida et al. / Journal of Analytical and Applied Pyrolysis xxx (2016) xxx–xxx thermal catalytic cracking of scum from grease traps in pilot scale, J. Anal. Appl. Pyrolysis 8 (January) (2016) (Available online). Başak Burcu Uzun, Ayşe Eren Pütün, Ersan Pütün, Fast pyrolysis of soybean cake: product yields and compositions, Bioresour. Technol. 97 (Issue 4) (2006) 569–576. Sevgi Şensöz, Dilek Angın, Pyrolysis of safflower (Charthamus tinctorius L.) seed press cake: part 1. The effects of pyrolysis parameters on the product yields, Bioresour. Technol. 99 (Issue 13) (2008) 5492–5497. Başak Burcu Uzun, Ayşe Eren Pütün, Ersan Pütün, Composition of products obtained via fast pyrolysis of olive-oil residue: effect of pyrolysis temperature, J. Anal. Appl. Pyrolysis 79 (Issues 1–2) (2007), 147–15. L. Dandik, H.A. Aksoy, Effect of catalyst of used Oil Carried Out in a Fractionating Pyrolysis reactor, Renew. Energ. 16 (1999) 1007–1010. V.R. Wiggers, A. Wisniewski Jr., L.A.S. Madureira, A.A. ChivangaBarros, H.F. Meier, Meier Biofuels from waste fish oil pyrolysis: continuous production in a pilot plant, Fuel 88 (Issue 11) (2009) 2135–2141. Takwa Kraiem, Aïda Ben Hassen-Trabelsi, Slim Naoui, Habib Belayouni, Mejdi Jeguirim, Characterization of the liquid products obtained from Tunisian waste fish fats using the pyrolysis process, Fuel Process. Technol. 138 (2015) 404–412. H.F. Meier, V.R. Wiggers, G.R. Zonta, D.R. Scharf, E.L. Simionatto, L. Ender, A kinetic model for thermal cracking of waste cooking oil based on chemical lumps, Fuel 144 (2015) 50–59. V.R. Wiggers, G.R. Zonta, A.P. França, D.R. Scharf, E.L. Simionatto, L. Ender, H.F. Meier, Challenges associated with choosing operational conditions for triglyceride thermal cracking aiming to improve biofuel quality, Fuel 107 (May) (2013) 601–608. Andre L.F. Santos, Danilo U. Martins, Osvaldo K. Iha, Rafael A.M. Ribeiro, Rafael L. Quirino, Paulo A.Z. Suarez, Agro-industrial residues as low-price feedstock for diesel-like fuel production by thermal cracking, Bioresour. Technol. 101 (Issue 15) (2010) 6157–6162 (August). E. Buzetzki, K. Sidorová, Z. Cvengrošová, J. Cvengroš, Effects of oil type on products obtained by cracking of oils and fats, Fuel Process. Technol. 92 (2011) 2041–2047. L. Dandik, H.A. Aksoy, Pyrolysis of used sunflower oil in the presence of sodium carbonate by using fractionating pyrolysis reactor, Fuel Process. Technol. 57 (2) (1998) 81–92. Y.S. Ooi, Ridzuan Zakaria, Abdul Rahman Mohamed, Subhash Bhatia, Catalytic conversion of palm oil-based fatty acid mixture to liquid fuel, Biomass Bioenergy 27 (Issue 5) (2004) 477–484. Y.S. Ooi, R. Zakaria, A.R. Mohamed, S. Bhatia, Synthesis of composite material MCM-41/beta and its catalytic performance in waste used palm oil cracking, Appl. Cata. A: Gen. 274 (Issues 1–2) (2004) 15–23. Y.S. Ooi, R. Zakaria, A.R. Mohamed, S. Bhatia, Catalytic conversion of fatty acids mixture to liquid fuels over mesoporous materials, React. Kinet. Catal. Lett. 84 (Issue 2) (2005) 295–302. Y.S. Ooi, R. Zakaria, A.R. Mohamed, S. Bhatia, Catalytic cracking of used palm oil and palm oil fatty acids mixture for the production of liquid fuel: kinetic modeling, J. Am. Chem. Soc. 18 (2014) 1555–1561. Y.S. Ooi, R. Zakaria, A.R. Mohamed, S. Bhatia, Catalytic conversion of fatty acids mixture to liquid fuel and chemicals over composite Microporous/Mesoporous catalysts, Energy Fuels 19 (3) (2005) 736–743. W.H. Chang, C.T. Tye, Catalytic cracking of used palm oil using composite zeolite, MJAS 17 (No. 1) (2013) 176–184. Witchakorn Charusiri and Tharapong Vitidsant, Kinetic study of used vegetable oil to liquid fuels over sulfated zirconia, Energy Fuels 19 (5) (2005) 1783–1789. B. Weber, E.A. Stadlbauer, S. Stengl, M. Hossain, A. Frank, D. Steffens, Elmar Schlich, Gerhard Schilling, Production of hydrocarbons from fatty acids and animal fat in the presence of water and sodium carbonate: reactor performance and fuel properties, Fuel 94 (2012) 262–269. E.A. Stadlbauer, R. Altensen, S. Bojanowski, G. Donnevert, A. Fiedler, S. Hossain, J. Rossmanith, G. Shilling, Herstellung von kohlenwasserstoffen aus tierfett durch thermokatalytisches spalten, Erdöl Erdgas Kohle 122 (2) (2006) 64–69. B. Weber, E.A. Stadlbauer, S. Eichenauer, A. Frank, D. Steffens, S. Elmar, S. Gerhard, Characterization of alkanes and olefins from thermo-chemical conversion of animal fat, J. Biobased Mater. Bioenergy 8 (No. 5) (2014) 526–537 (12). [39] M.Y. Liew, A. Salmiaton, W.A.K.G. Wan Azlina, R. Yunus, Catalytic pyrolysis of waste chicken fats using zeolite catalysts, in: Developments in Sustainable Chemical and Bioprocess Technology, Springer, US, 2013, ISBN 978-1-4614-6208-8, pp. 73–79, http://dx.doi.org/10.1007/978-1-4614-62088 10. [40] Hua Tian, Chunyi Li, Chaohe Yang, Honghong Shan, Alternative processing technology for converting vegetable oils and animal fats to clean fuels and light olefins, Chin. J. Chem. Eng. 16 (Issue 3) (2008) 394–400. [41] S. Eichenauer, B. Weber, E.A. Stadlbauer, Thermochemical processing of animal fat and meat and bone meal to hydrocarbons based fuels, ASME 2015, 9th International Conference on Energy Sustainability (2015) (Paper No. ES2015-49197, V001T02A001, 10.115/ES2015-49197). [42] S. Bojanowski, A. Fiedler, A. Frank, E.A. Stadlbauer, G. Schilling, Catalytic production of liquid fuels from organic residues of rendering plants, Eng. Life Sci. 7 (No. 6) (2007) 593–598. [43] Ooi Yean Sang, Biofuel production from catalytic cracking of palm oil, Energy Sources 25 (Issue 9) (2003) 859–869. [44] Farouq A. Twaiq, Abdul Rahman Mohamed, Subhash Bhatia, Liquid hydrocarbon fuels from palm oil by catalytic cracking over aluminosilicate mesoporous catalysts with various Si/Al ratios, Microporous Mesoporous Mater. 64 (Issues 1–3) (2003) 95–107. [45] Fadouq A. Twaiq, Noor A.M. Zabidi, Subhash Bhatia, Catalytic conversion of palm oil to hydrocarbons: performance of various zeolite catalysts, Ind. Eng. Chem. Res. 38 (9) (1999) 3230–3237. [46] A.A.Twaiq Farouq, A.R. Mohamad, Subhash Bhatia, Performance of composite catalysts in palm oil cracking for the production of liquid fuels and chemicals, Fuel Process. Technol. 85 (Issue 11) (2004) 1283–1300. [47] A. Farouq, Noor Asmawati M. Zabidi, Abdul Rahman Mohamad, Subhash Bhatia, Catalytic conversion of palm oil over mesoporous aluminosilicate MCM-41 for the production of liquid hydrocarbon fuels, Fuel Process. Technol. 84 (Issues 1–3) (2003) 105–120. [48] Dessy Y. Siswanto, Giyanto W. Salim, Nico Wibisono, Herman Hindarso, Yohanes Sudaryanto, Suryadi Ismadji, Gasoline production from palm oil via catalytic cracking using MCM41: determination of optimum conditions, J. Eng. Appl. Sci. 3 (No. 6) (2008) 42–46. [49] E. Buzetzki, K. Sidorová, Z. Cvengrošová, A. Kaszonyl, J. Cvengroš, The influence of zeolite catalysts on the products of rapeseed oil craking effects of oil type on products obtained by cracking of oils and fats, Fuel Process. Technol. 92 (2011) 1623–1631. [50] Daniela G. Lima, Valerio C.D. Soares, Eric B. Ribeiro, Daniel A. Carvalho, Érika C.V. Cardoso, Flávia C. Rassi, Kleber C. Mundim, Joel C. Rubim, A.Z. Paulo, Suarez D.G. Lima, V.C.D. Soares, E.B. Ribeiro, D.A. Carvalho, E.C.V. Cardoso, F.C. Rassi, Diesel-like fuel obtained by pyrolysis of vegetable oils, J. Anal. Appl. Pyrolysis 71 (2004) 987–996. [51] V.R. Wiggers, H.F. Meier, A. Wisniewski Jr., A.A. ChivangaBarros, M.R. WolfMaciel, Biofuels from continuous fast pyrolysis of soybean oil: a pilot plant study, Bioresour. Technol. 100 (Issue 24) (2009) 6570–6577. [52] S.A.P. Mota, A.A. Mancio, D.E.L. Lhamas, D.H. de Abreu, M.S. da Silva, W.G. dos Santos, D.A.R. deCastro, R.M. deOliveira, M.E. Araújo, L.E.P. Borges, N.T. Machado, Production of green diesel by thermal catalytic cracking of crude palm oil (Elaeis guineensis Jacq) in a pilot plant, J. Anal. Appl. Pyrolysis 110 (2014) 1–11. [53] ANP No. 65, 9.12.2011 – DOU 12.12.2011. [54] Thermokatalytische Niedertemperaturkonvertierung (NTK) von tierischer und mikrobieller Biomasse unter Gewinnung von Wertstoffen und Energieträgern im Pilotmaßstab. Abschlussbericht. Band 1 (Hauptteil). Entwicklungsprojekt gefördert unter dem Az. 18153 von der Deutschen Bundesstiftung Umwelt, Osnabrück. https://www.dbu.de/OPAC/ab/DBUAbschlussbericht-AZ-18153.pdf. [55] A. Kubátová, Y. Luo, J. Št’ávová, S.M. Sadrameli, T. Aulich, E. Kozliak, W. Seames, New path in the thermal cracking of triacylglycerols (canola and soybean oil), Fuel 90 (2011) 2598–2608. [56] X. Dupain, D.J. Costa, C.J. Schaverien, M. Makkee, J.A. Moulijn, Cracking of a rapeseed vegetable oil under realistic FCC conditions, Appl. Catal. B 72 (2007) 44–61. [57] ANP No. 37, 1.12.2009 – DOU 02.12.2009. [58] W.L. Nelson, Petroleum Refinery Engineering, 4th edition, McGraw-Hill, 1985. Please cite this article in press as: H. da Silva Almeida, et al., Performance of thermochemical conversion of fat, oils, and grease into kerosene-like hydrocarbons in different production scales, J. Anal. Appl. Pyrol. (2016), http://dx.doi.org/10.1016/j.jaap.2016.04.017