Format And Type Fonts CCHHEEMMIICCAALL EENNGGIINNEEEERRIINNGG TTRRAANNSSAACCTTIIOONNSS VOL. 45, 2015 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Sharifah Rafidah Wan Alwi, Jun Yow Yong, Xia Liu Copyright © 2015, AIDIC Servizi S.r.l., ISBN 978-88-95608-36-5; ISSN 2283-9216 DOI: 10.3303/CET1545222 Please cite this article as: Nyakuma B.B., Ahmad A., Johari A., Tuan Abdullah T.A., Oladokun O., Aminu D.Y., 2015, Non- isothermal kinetic analysis of oil palm empty fruit bunch pellets by thermogravimetric analysis, Chemical Engineering Transactions, 45, 1327-1332 DOI:10.3303/CET1545222 1327 Non-Isothermal Kinetic Analysis of Oil Palm Empty Fruit Bunch Pellets by Thermogravimetric Analysis Bemgba B. Nyakuma* ,a , Arshad Ahmad a , Anwar Johari a , Tuan A. Tuan Abdullah a , Olagoke Oladokun a , Dodo Y. Aminu b a Institute of Future Energy, Centre for Hydrogen Energy, Universiti Teknologi Malaysia, 81310 Skudai, Johor Bahru Malaysia. b Department of Architecture, Faculty of Built Environment, Universiti Teknologi Malaysia, 81310 Skudai, Johor Bahru, Malaysia. bbnyax1@gmail.com The pyrolysis kinetics of oil palm empty fruit bunch (OPEFB) pellets was examined under non-isothermal conditions in a thermogravimetric (TG) analyser. Thermal analysis was carried out from 30 °C to 1,000 °C using three different heating rates 5, 10, 20 °C min -1 under nitrogen gas (N2). The TG-DTG curves showed that the pyrolysis process occurred in three steps; drying, active pyrolysis and passive pyrolysis signifying the removal of moisture, holocellulose and lignin. The pyrolysis kinetic parameters; activation energy, Ea, and frequency factor A, were deduced from the Flynn-Wall-Ozawa (FWO) model. The average Ea and A values from α = 0.10 - 0.60 were 160.20 kJ/mol and 1.38 x 10 24 min -1 . The highest Ea (231.42 kJ/mol) and A (8.27 x 10 24 min -1 ) occurred at α = 0.30 indicating this is the slowest or rate determining step (RDS) during thermal degradation of OPEFB pellets. The average Ea for OPEFB pellets was comparably lower than cornstalk (206.40 kJ/mol), sawdust (232.60 kJ/mol) and oak (236.20 kJ/mol). The kinetic compensation or isokinetic effect was also observed during thermal decomposition of the OPEFB pellets. Hence, the results indicate OPEFB pellets can be utilized as a potential feedstock for pyrolysis. 1. Introduction The large scale cultivation of oil palm (Elaeis guineensis) in Malaysia generates enormous quantities of biomass waste such as empty fruit bunch (EFB). Consequently, the inefficiency of current conversion technologies for oil palm waste (OPW) has created socioeconomic, environmental and technological challenges in Malaysia. Therefore, the valorization of OPW into fine chemicals, power generation and clean energy fuels will address the challenges of OPW accumulation in the oil palm industry. Furthermore, the utilization of biomass can potentially address the problems of price volatility, dwindling reserves and environmental pollution associated with fossil fuels (Johari et al., 2015). Conversely, biomass is mainly available as high moisture, inhomogeneous sized and low energy density fuels which require pre-treatment and conditioning before utilization (Kuparinen et al., 2014). The most widely used biomass pre-treatment techniques include drying, pelletization (Nyakuma et al., 2014a) and torrefaction (Basu, 2010). These techniques can improve the feedstock supply chain (Eranki et al., 2011), thermochemical properties (Nyakuma et al., 2014c) and potential applications of OPW (Kelly-Yong et al., 2007) in thermal bioenergy conversion systems (BCS). Consequently, research groups in Malaysia are currently investigating efficient conversion techniques for valorising OPW via torrefaction (Uemura et al., 2011), gasification (Lahijani and Zainal, 2014) for clean energy applications (Aziz et al., 2015). The transition from fossil fuels to clean bioenergy fuels requires fundamental knowledge of thermal behaviour and decomposition kinetics of biomass (Munir et al., 2009). This is vital for the feasibility, design and scaling up biomass thermal conversion equipment (Ma et al., 2012). Furthermore, the kinetics of biomass decomposition can be useful in optimizing the yield and composition of desired products during thermochemical conversion (Islam et al., 2015). Consequently, the mathematical models developed by Flynn-Wall (Flynn and Wall, 1966) and Ozawa (Ozawa, 1965) along with analytical techniques such as mailto:bbnyax1@gmail.com 1328 thermogravimetric analysis (TGA) have been successfully applied to investigate the decomposition kinetics of biomass (Mortari et al., 2014). To the best of the authors’ knowledge, comprehensive studies on the effects of pelletization on the thermal degradation behaviour and decomposition kinetics of OPW is lacking in literature. This study is aimed at investigating the thermal degradation behaviour of oil palm empty fruit bunches (OPEFB) pellets under non-isothermal conditions using thermogravimetric analyser (TGA). Consequently, the decomposition kinetics of the OPEFB pellets during pyrolytic TG analysis is presented using the Flynn-Wall-Ozawa (FWO) kinetic model. 2. Experimental 2.1 Ultimate and Proximate Analysis The oil palm empty fruit bunches (OPEFB) pellets was supplied by an oil palm mill in Johor, Malaysia. The OPEFB pellets were subsequently pulverised and sifted to obtain 125 µm sized particles. The elemental composition and proximate analysis was examined in as received (ar) basis using standard ASTM techniques, as presented in Table 1. A detailed characterization of the thermochemical fuel properties of OPEFB pellets in presented in our previous study (Nyakuma et al., 2014b). Table 1: Ultimate and Proximate analysis of OPEFB pellets C H N S O M VM FC A HHV 45.14 6.05 0.54 0.20 48.08 8.11 72.1 14.91 4.89 17.57 C-Carbon, H-Hydrogen, N-Nitrogen, S-Sulphur, O-Oxygen, M-Moisture, VM-Volatile matter, FC-Fixed carbon, A-Ash, HHV-Higher heating value in MJ/kg. 2.2 Thermal analysis Thermal analysis of OPEFB was carried out in a thermogravimetric (TG) analyser (Netzsch TM 209 F3) under nitrogen flow rate of 50 mL min -1 . During each run, the temperature program was set to heat the sample from 30 ºC to 1,000 ºC using three heating rates, β = 5, 10, 20 ºC min -1 . Subsequently, the FWO model was applied to determine decomposition kinetic parameters; activation energy, Ea, and frequency factor, A at different conversions, α, during TG analysis. For thermally degrading biomass, the rate of solid state decomposition can be expressed as; ( ) ( ) d k T f dt   (1) Where f(α) is the reaction model, and k(T) - temperature dependent rate constant given by; ( ) A exp Ea k T RT        (2) Where A is the frequency factor, Ea - activation energy, R - universal gas constant, and T - absolute temperature. Relating Eqs(1) and (2) yields the expression for temperature T, dependence on conversion, α, exp ( ) d Ea A f dt RT          (3) By taking into account the effect of heating rate, β, the integral conversion function, g(α) can be deduced; 0 0 ( ) exp ( ) Td A Ea g dT f RT                (4) The expression in Eq(4) is the fundamental equation for analysing the decomposition kinetic parameters of thermal biomass conversion. In this study, the model free isoconversional Flynn-Wall-Ozawa method, derived from Doyle’s approximation (Doyle, 1965), was selected to analyse decomposition kinetics of OPEFB pellets. It can be expressed as; 1329 In( ) In 5.331 1.052 ( ) AEa Ea Rg RT                (5) Hence, from the plot of In(β) against (1/T) at different heating rates, the activation energy can be deduced from the slope –Ea/R where the value of R is 8.314 J/mol K and frequency factor by In[AR/Ea)]. 3. Results and Discussion 3.1 Thermal analysis The TG and DTG curves for OPEFB pellets are presented in Figures 1 and 2. The plots display the reverse S-curves typically observed for thermal degradation of biomass during TG analysis. In addition, the plots showed that increase in heating rate from 5 – 20 °C min -1 led to a shift in the TG-DTG curves to higher temperatures signifying the temperature dependency of the pyrolytic thermal decomposition process (Slopiecka et al., 2012). The temperature-heating rate effect can be aptly examined from the characteristic decomposition profiles presented in Table 2. Table 2: Characteristic temperatures for OPEFB pellet decomposition at different heating rates Heating Rate (°C/min) Reaction Time (min) Peak Temperature (°C ) Residual Weight (%) 5 194.00 308.30 11.03 10 97.00 316.70 15.29 20 48.50 329.60 22.02 From Table 2, it is evident that varying heating rate resulted in a corresponding increase in peak decomposition temperature and residual weight of OPEFB pellets. This is reportedly due a heating transfer limitation which ensures that the thermal energy required for complete pyrolytic decomposition decreases at higher heating rates (Slopiecka et al., 2012). This eventually results in an increase in peak temperatures and residual mass as reported for OPEFB pellets in Table 2. Figure 1: TG curves of Oil Palm Empty Fruit Bunch (OPEFB) pellets. Figure 2: DTG curves of Oil Palm Empty Fruit Bunch (OPEFB) pellets From Figures 1 and 2, the pyrolytic decomposition of the OPEFB pellets clearly occurred in 3 distinct stages; drying (50 - 200 °C), active pyrolysis (200 – 500 °C) and passive pyrolysis (500 – 1,000 °C). The drying state is characterised by the removal of surface moisture water (Açıkalın, 2011) and low molecular weight volatiles (Çepelioğullar and Pütün, 2014). The active pyrolysis stage is characterized by significant mass loss due to the degradation of the biomass components hemicellulose and cellulose. The passive pyrolysis stage is categorised by slow mass loss rate attributed to lignin degradation (Souza et al., 2009) as denoted by the “tailing” observed above 500 °C in Figure 2. 1330 3.2 Kinetic Analysis The linear regression plots of In β vs 1/T for conversions α = 0.10 – 0.60 for OPEFB pellets are presented in Figure 3. The slope is given by -1.052 Ea/R while frequency factor A was deduced from the intercept of the plots using the relation In (AR/Ea) based on the assumption 2RT << Ea (Damartzis et al., 2011). Furthermore, the conversions α < 0.1 and α > 0.8 have been excluded due to the low correlation values (Damartzis et al., 2011). Figure 3: Kinetic plot for OPEFB pellets using the Flynn-Wall-Ozawa (FWO) model The kinetic plots observed for conversions α = 0.2 to 0.6 in Figure 3 indicate that the OPEFB decomposition mechanism is characterized by parallel reactions occurring simultaneously. Furthermore, the kinetic parameters and the mechanism of conversion is significantly influenced by complex reactions, typified by the fluctuating Ea and A values (Ceylan and Topçu, 2014). The slope and intercept of the plots were used to calculate kinetic parameters for OPEFB pellets decomposition for conversions α = 0.1 - 0.6 as presented in Table 3. Table 3: Calculated kinetic parameters for OPEFB pellets Conversion (α) R 2 E (kJ/mol) A (min -1 ) 0.10 0.9985 73.42 1.80 x 10 10 0.20 0.9899 159.10 2.50 x 10 18 0.30 0.9975 231.42 8.27 x 10 24 0.40 0.9898 183.37 4.39 x 10 19 0.50 0.9986 154.56 3.35 x 10 16 0.60 0.9638 159.37 3.50 x 10 16 Average 0.9897 160.21 1.38 x 10 24 The activation energy, Ea increased from 73.42 to 231.42 kJ/mol while and frequency factor, A was from 1.80 x 10 10 to 8.27 x 10 24 min -1 with average correlation R 2 = 0.99. Comparatively, the average Ea for EFB pellets (160.21 kJ/mol) is lower than cornstalk (206.40 kJ/mol), sawdust (232.60 kJ/mol) and oak tree (236.20 kJ/mol) in literature (Sun et al., 2012). Since activation energy is the minimum energy requirement for reactants before the start of a chemical reaction, high Ea values will result in slower reactions. Therefore, the lower Ea values of OPEFB pellets emphasizes its suitability as a feedstock for pyrolysis. 1331 3.3 Kinetic Compensation Effect (KCE) The interdependency of the kinetic parameters Ea and A are presented in Figure 4. This linear relationship between the kinetic parameters is termed the kinetic compensation effect (KCE) (Slopiecka et al., 2012) or the isokinetic effect (Açıkalın, 2011). Figure 4: Interdependency of apparent activation energy Ea and Frequency factor A for OPEFB pellets. From Figure 4, the highest Ea and A values during kinetic analysis were observed at α = 0.30 between α = 0.20 – 0.40. The activation energy at α = 0.30 was 231.42 kJ/mol which is significantly higher than the average apparent activation energy (Ea = 160.20 kJ/mol) of the entire OPEFB pellets pyrolysis process. This indicates that reaction is slowest at this stage and requires a high energy of activation and collision between the reacting particles to proceed to completion. Therefore, this can be denoted as the rate determining step (RDS) as also evidenced by the significant mass loss observed during thermal degradation. 4. Conclusions The pyrolysis kinetics of OPEFB pellets pyrolysis was investigated under non-isothermal conditions using a TG analyser. The TG/DTG (mass loss) curves indicated that pyrolysis occurred in three stages; drying, active pyrolysis and passive pyrolysis. The Flynn-Wall-Ozawa (FWO) model was applied to the TG/DTG data to deduce the kinetic parameters, activation energy, Ea, and frequency factor A. The average Ea and A values were 160.20 kJ/mol and 1.38 x 10 24 min -1 . Furthermore, kinetic analysis also indicated the effect of kinetic compensation during thermal decomposition of the OPEFB pellets. In addition, the average Ea value of the OPEFB pellets was lower than other biomass waste such as cornstalk, sawdust, and oak tree reported which confirms that OPEFB pellets is a potentially suitable feedstock for biomass pyrolysis. Acknowledgment The authors acknowledge the Ministry of Education (MOE) Malaysia for the Long Research Grant Scheme (LRGS) VOT: 4L817. The authors are grateful to S. L. Wong, and Muhamad Faizal B. A. Halim of Universiti Teknologi MARA, Shah Alam for the TG measurements. References Açıkalın K., 2011, Thermogravimetric analysis of walnut shell as pyrolysis feedstock, Journal of Thermal Analysis and Calorimetry, 105(1), 145-150. Aziz M., Prawisudha P., Prabowo B., Budiman B.A., 2015, Integration of energy-efficient empty fruit bunch drying with gasification/combined cycle systems, Applied Energy, 139, 188-195. 1332 Basu P., 2010, Biomass gasification and pyrolysis: practical design and theory. Academic Press, Burlington, MA, USA. Çepelioğullar Ö., Pütün A.E., 2014, A pyrolysis study for the thermal and kinetic characteristics of an agricultural waste with two different plastic wastes, Waste Management and Research, 32(10), 971- 979. Ceylan S., Topçu Y., 2014, Pyrolysis kinetics of hazelnut husk using thermogravimetric analysis, Bioresource Technology, 156, 182-188. Damartzis T., Vamvuka D., Sfakiotakis S., Zabaniotou A., 2011, Thermal degradation studies and kinetic modeling of cardoon (Cynara cardunculus) pyrolysis using thermogravimetric analysis (TGA), Bioresource Technology, 102(10), 6230-6238. Doyle C.D., 1965, Series Approximations to the Equation of Thermogravimetric Data, Nature, 207(4994), 290-291. Eranki P.L., Bals B.D., Dale B.E., 2011, Advanced Regional Biomass Processing Depots: a key to the logistical challenges of the cellulosic biofuel industry, Biofuels, Bioproducts and Biorefining, 5(6), 621- 630. Flynn J.H., Wall L.A., 1966, A quick, direct method for the determination of activation energy from thermogravimetric data, Journal of Polymer Science Part B: Polymer Letters, 4(5), 323-328. Islam M.A., Asif M., Hameed B.H., 2015, Pyrolysis kinetics of raw and hydrothermally carbonized Karanj (Pongamia pinnata) fruit hulls via thermogravimetric analysis, Bioresource Technology, 179, 227-233. Johari A., Nyakuma B.B., Mohd Nor S.H., Mat R., Hashim H., Ahmad A., Yamani Zakaria Z., Tuan Abdullah T.A., 2015, The challenges and prospects of palm oil based biodiesel in Malaysia, Energy, 81, 255-261. Kelly-Yong T.L., Lee K.T., Mohamed A.R., Bhatia S., 2007, Potential of hydrogen from oil palm biomass as a source of renewable energy worldwide. Energy Policy, 35(11), 5692-5701. Kuparinen K., Heinimö J., Vakkilainen E., 2014, World's largest biofuel and pellet plants – geographic distribution, capacity share, and feedstock supply, Biofuels, Bioproducts and Biorefining, 8(6), 747-754. Lahijani P., Zainal Z.A., 2014, Fluidized Bed Gasification of Palm Empty Fruit Bunch Using Various Bed Materials, Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 36(22), 2502- 2510. Ma Z., Zhang Y., Zhang Q., Qu Y., Zhou J., Qin H., 2012, Design and experimental investigation of a 190 kWe biomass fixed bed gasification and polygeneration pilot plant using a double air stage downdraft approach, Energy, 46(1), 140-147. Mortari D.A., Britto M.C., Crnkovic P.M., 2014, Correlation Between Activation Energy and Thermal Decomposition Yield of Sugar cane Bagasse under CO2/O2 and N2/O2, Chemical Engineering Transactions, 37(31-36). Munir S., Daood S., Nimmo W., Cunliffe A., Gibbs B., 2009, Thermal analysis and devolatilization kinetics of cotton stalk, sugar cane bagasse and shea meal under nitrogen and air atmospheres, Bioresource Technology, 100(3), 1413-1418. Nyakuma B.B., Johari A., Ahmad, A., Abdullah, T.A.T., 2014a, Comparative analysis of the calorific fuel properties of Empty Fruit Bunch Fiber and Briquette, Energy Procedia, 52, 466-473. Nyakuma B.B., Mazangi M., Tuan Abdullah T.A., Johari A., Ahmad A., Oladokun O., 2014b, Gasification of Empty Fruit Bunch Briquettes in a Fixed Bed Tubular Reactor for Hydrogen Production, Applied Mechanics and Materials, 699, 534-539. Nyakuma B.B., Oladokun O.A., Johari A., Ahmad A., Abdullah T.A.T., 2014c, A Simplified Model for Gasification of Oil Palm Empty Fruit Bunch Briquettes, Jurnal Teknologi, 69(2). Ozawa T., 1965, A New Method of Analyzing Thermogravimetric Data, Bulletin of the Chemical Society of Japan, 38(11), 1881-1886. Slopiecka K., Bartocci P., Fantozzi F., 2012, Thermogravimetric analysis and kinetic study of poplar wood pyrolysis, Applied Energy, 97, 491-497. Souza B., Moreira A., Teixeira A.F., 2009, TG-FTIR coupling to monitor the pyrolysis products from agricultural residues, Journal of Thermal Analysis and Calorimetry, 97(2), 637-642. Sun W.G., Zhao H., Yan H.X., Sun B.B., Dong S.S., Zhang C.W., Qin S., 2012, The Pyrolysis Characteristics and Kinetics of Jerusalem artichoke Stalk Using Thermogravimetric Analysis, Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 34(7), 626-635. Uemura Y., Omar W.N., Tsutsui T., Yusup S.B., 2011, Torrefaction of oil palm wastes, Fuel, 90(8), 2585- 2591.