Investigation on Thermochemical Behaviour of Low Rank Malaysian Coal, Oil Palm Biomass and Their Blends During Pyrolysis via Thermogravimetric Analysis (TGA)

Bioresource Technology 101 (2010) 4584–4592 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech Investigation on thermochemical behaviour of low rank Malaysian coal, oil palm biomass and their blends during pyrolysis via thermogravimetric analysis (TGA) Siti Shawalliah Idris a,*, Norazah Abd Rahman a, Khudzir Ismail b, Azil Bahari Alias a, Zulkifli Abd Rashid a, Mohd Jindra Aris a a b Faculty of Chemical Engineering, Universiti Teknologi MARA Malaysia, 40450 Shah Alam, Selangor, Malaysia Faculty of Applied Sciences, Universiti Teknologi MARA Perlis, 02600 Arau, Perlis, Malaysia a r t i c l e i n f o Article history: Received 16 September 2009 Received in revised form 12 January 2010 Accepted 14 January 2010 Available online 12 February 2010 Keywords: Devolatilisation Oil palm wastes Sub-bituminous coal Non-isothermal TG analysis Coal/biomass blends a b s t r a c t This study aims to investigate the behaviour of Malaysian sub-bituminous coal (Mukah Balingian), oil palm biomass (empty fruit bunches (EFB), kernel shell (PKS) and mesocarp fibre (PMF)) and their respective blends during pyrolysis using thermogravimetric analysis (TGA). The coal/palm biomass blends were prepared at six different weight ratios and experiments were carried out under dynamic conditions using nitrogen as inert gas at various heating rates to ramp the temperature from 25 °C to 900 °C. The derivative thermogravimetric (DTG) results show that thermal decomposition of EFB, PMF and PKS exhibit one, two and three distinct evolution profiles, respectively. Apparently, the thermal profiles of the coal/oil palm biomass blends appear to correlate with the percentage of biomass added in the blends, thus, suggesting lack of interaction between the coal and palm biomass. First-order reaction model were used to determine the kinetics parameters for the pyrolysis of coal, palm biomass and their respective blends. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction Renewable energy has become more important globally especially with the current fuel and economic crisis. In Malaysia, the government has encouraged the use of renewable energy through its Five Fuels Policy plan in 1999 with the estimation of 5% utilisation of renewable energy in the energy mix for year 2008 and this usage will increase to 35% in 2030 (Bernama, 2008). Malaysia is well positioned amongst the ASEAN countries to promote the use of biomass as a renewable energy source in her national energy mix since she is a major agricultural commodity producer in the region. The favourable climate conditions that prevail throughout the year are an advantage for palm oil cultivation. Evidently, Malaysia is the world’s largest producer and exporter of palm oil, replacing Nigeria as the chief producer since 1971 (Yusoff, 2006). At present, there are more than 3.88 million hectares of land under oil palm cultivation and it is the main contributor to biomass resources in Malaysia. Approximately 368 palm oil mills are operating in the country to date, thus producing substantial amount of lignocellulosic biomass in the form of empty fruit bunches (EFB), palm kernel shell (PKS) and palm mesocarp fibre (PMF). It was estimated that the amount of solid wastes produced could reach 39 * Corresponding author. Tel.: +60 3 55436312; fax: +60 3 554356300. E-mail address: shawal075@salam.uitm.edu.my (S.S. Idris). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.01.059 million tons by the year 2020 (Yusoff, 2006). Even though there are a number of researches investigating the ways of utilising these wastes there is still a large fraction of the residues abundantly unattended (Yusoff, 2006). Currently, most of the wastes are incinerated or utilised as boiler fuels for steam generation to the palm oil mills. They have also been used as mulching around the plants so as to prevent excessive evaporation or erosion, producing medium density fibre board in furniture and used as material for mattresses, seats, insulation and paper making industries (Chuah et al., 2006; Kalam et al., 2004; Kelly-Yong et al., 2007). Oil palm biomasses are highly potential materials for energy resources. The fact that they are renewable and abundantly available are amongst the attractive reasons of employing them as the major source for renewable energy (Yusoff, 2006). Furthermore, it appears to have soundly positive environmental properties resulting in net zero releases of carbon dioxides and very low sulphur content. In practice, about half of the agricultural residues are consumed for energy generation. This amount contributes to about 20% of the primary energy demand of industries in Malaysia. The role of biomass is presently limited in power development, but opportunities exist for increasing its share. It is estimated that by 2050 biomass could provide nearly 38% of the world’s direct fuel use and 17% of the world’s electricity (Bernama, 2008). Even so, with respect to power production, it appears that, a total replacement of coal for biomass requires further development. 4585 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Combustion of fossil fuels such as coal contributes to CO2 emission that causes global warming effect to the environment. A means of reducing the CO2 emissions is by minimizing coal combustion by implementing coal-biomass co-combustion and it is one of the promising short term alternatives for the use of renewable fuels. Moreover, the use of biomass may represent an alternative to the disposal of waste and its exploitation as energy source. Coal availability worldwide and price stability, are essential components in the long term for continuing utilisation around the world. In Malaysia, coal reserve stands at about 1712 million tones of various ranks ranging from lignite to anthracite. Apparently, Malaysia has moved towards utilising coal through National Mineral Policy for power production. On a long term, coal will have a greater contribution in Malaysia’s energy mix at about 40–45% in comparison to that of natural gas with contribution of less than 50% (Rahman Mohamed and Lee, 2006). Pyrolysis of coal, biomass and coal/biomass blends are relatively new area of research and several researchers have investigated their pyrolysis and combustion behaviour (Caballero et al., 1997a; Caballero et al., 1997b; Ismail et al., 2005b; Kastanaki et al., 2002; Pan et al., 1996). Recently, Vamvuka et al. (2003) developed a kinetic modelling for the volatile matter released during the pyrolysis of several biomass (i.e. olive kernel, forest and cotton residues) blends with lignite using thermogravimetry (TG). Their findings revealed that the biomass possess higher thermochemical reactivity with shorter devolatilisation times in comparison to the lignite. In another work, Moghtaderi et al. (2004) investigated the pyrolytic behaviour of coal/woody biomass (i.e. pine dust) blends over a wide range of heating rates and temperatures using tubular reactors in an attempt to simulate the pulverized fuel boilers. Most important, they found that the total yield of the major pyrolysis products were linearly proportional to the blending ratio, indicating no synergistic effect between coal and biomass. Likewise, Vuthaluru (2004) also reported the same observation during the investigation on the thermal behaviour during co-pyrolysis of coal/biomass (i.e. wood waste and wheat straw) using thermogravimetry. Up to date, the study of Malaysian coal and oil palm biomass copyrolysis has not yet been reported. Nevertheless, few studies pertaining to pyrolysis of oil palm biomass were presented. Yang et al. (2004) had investigated the pyrolysis of oil palm biomass namely PMF, PKS, and EFB using TGA-FTIR analyser. The investigation was carried out to study the effects of particle size and heating rates on the evolution of gases during pyrolysis process. From this study, they found out that the particle size between 250 lm and 2 mm has no significant influence on pyrolysis process and that the pyrolysis kinetics reaction for the palm biomass is of first-order reaction. Luangkiattikhun et al. (2008) in their work, had modelled the kinetics of pyrolysis of palm kernel, palm mesocarp fibre and palm kernel shell using non-isothermal thermogravimetric method. They proposed two parallel reaction model to describe pyrolysis reaction of palm kernel shell and mesocarp fibre while one step global kinetic to describe pyrolysis of palm kernel. In the development of converting coal-biomass blends into renewable energy sources, understanding of the chemical composition, thermal behaviour and reactivity of oil palm biomass and their blends with coal during pyrolysis, is of paramount important. This is because pyrolysis or solid devolatilisation is always a first step in the thermochemical conversion studies towards producing this renewable energy sources. Therefore the aim of this work is to investigate the thermal behaviour and reactivity of Mukah Balingian coal, PMF, PKS, EFB, coal/PMF, coal/PKS and coal/EFB blends during pyrolysis process using thermogravimetry analyser. The thermogravimetric technique was used to identify the thermal evolution profiles occurred during pyrolysis and to obtain the kinetic parameter of these materials during the process. The effect of varying heating rates on the reactivity of coal, PMF, PKS, and EFB and their respective blends on the kinetic parameters will also be reported. 2. Methods 2.1. Materials Malaysian low rank coal (Mukah Balingian) originated from Sarawak and oil palm biomass namely EFB, PMF and PKS obtained from oil palm mill located in Nilai, Negeri Sembilan, were selected in this study. After air-drying for 2–3 days, the samples were crushed and sieved to desired particle size (<212 lm) using Retsch SM2000 heavy duty cutting mill and Endecott Shaker respectively. Finally, the samples were dried in a vacuum oven set at 80 °C for 24 h and kept in a tightly screw cap bottle. The characterisations of coal and biomass samples were shown in Table 1. Proximate analysis was carried out according to ASTM standards (ASTM D 5142 – 02a) using thermobalance TGA/SDRA51e manufactured by Mettler Toledo, while ultimate analysis was performed using a Thermo Finnigan Flashed 1112 analyzer according to ASTM D 5373 – 02. The calorific value was achieved based on ASTM D 2015 – 96 via Ika-works C5000 calorimeter. The analysis on mineral matter content in each samples were done using X-ray Fluorescence (XRF) model Bruker S4-Explorer X-ray Fluorescence (1 kW) and the results were reported in Table 2. 2.2. Thermogravimetry analysis A thermobalance TGA/SDRA51e was used for the pyrolysis experiments. Its precision of temperature measurement was ±0.5 K; microbalance sensitivity was less than 1.0 lg, while heating rate could be changed from 1 to 100 °C/min. Prior to the thermal investigation, the instrument is calibrated using indium/ aluminium check method. The two compounds of approximately 10 mg each were inserted in a 150 ll alumina crucible next to each other as such the two will not touch. The calibration was performed based on the set heating profiles from the manufacturer. The furnace and sample temperature was tested to be within the specified allowable specification. The pyrolysis investigations were carried out using nitrogen as carrier gas, under dynamic conditions at heating rates of 10, 20, 40 and 60 °C/min. The pyrolysis temperature was raised from room temperature to 900 °C. The biomass was added to coal at weight ratios of 0:100, 20:80, 40:60, 50:50, 60:40, 80:20 and 100:0. To avoid the problem of mass and heat transfer, a sample size of 20 mg was selected for this experiment. The weighted sample were inserted directly into 150 ll ceramic crucible and the temperature was kept isothermal for one minute Table 1 Ultimate and proximate analyses of Mukah Balingian coal, PMF, PKS and EFB. Analyses Coal Biomass Mukah Balingian EFB PMF PKS Proximate analysis (db) (wt.%) Volatile matter 42.2 Fixed carbon 48.8 Ash 5.8 70.5 15.4 4.5 68.8 15.2 10.2 69.2 16.0 10.5 Ultimate analysis (daf) (wt.%) Carbon 54.37 Hydrogen 5.29 Nitrogen 1.75 Sulphur 0.25 a 38.34 Oxygen Calorific value (MJ/kg) 24.6 40.93 5.42 1.56 0.31 51.78 16.8 43.19 5.24 1.59 0.19 49.79 19.0 41.33 4.57 0.99 0.09 53.02 16.3 db – Dry basis; daf – dry ash free basis. a Calculated by difference. 4586 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Table 2 XRF analysis for Mukah Balingian coal, PMF, PKS and EFB (wt.%). Samples Al Ca Cl Fe K Mg P S Si Sr Ti Y Zn MB coal EFB PMF PKS 16.60 0.45 1.96 0.84 9.77 15.40 14.60 58.96 1.10 8.83 5.19 0.00 24.82 4.84 38.56 33.71 2.89 61.53 18.00 1.98 0.63 1.40 0.38 0.19 3.76 1.48 2.20 0.67 18.10 1.41 2.66 0.38 17.90 3.96 16.00 2.51 0.52 0.04 0.00 0.11 2.51 0.17 1.15 0.37 0.81 0.00 0.00 0.00 0.64 0.22 0.43 0.10 until a steady condition was obtained before ramping to the desired temperature. Each individual samples was pyrolysed at least twice but more repetitions were carried out in case some variability was observed. 3. Results and discussion 3.1. Thermal decomposition of single fuel: coal, oil palm biomass The results of TGA analysis are shown in Fig. 1(a) and (b) which show the weight loss curves (TG) and derivative thermogravimetric (DTG) evolution profiles respectively, as a function of reaction temperature, for coal and oil palm biomass using heating rate of 10 °C/min. It is worth noting that the DTG evolution profile corresponds to the region where the slope of TG curve is constant (Ismail et al., 2005a; Vuthaluru, 2004). Thermal decomposition of 100 90 (a) Mukah Balingian Coal 80 (MB) Weight (%) 70 60 50 Palm Kernel Shell (PKS) Empty Fruit Brunch (EFB) 40 Palm Mesocarp Fibre (PMF) 30 20 25 100 200 300 400 500 600 700 800 900 Temperature (oC) 0.02 Derivative Weight (%/min) 0.01 (b) 0.0 - 0.01 Mukah Balingian Coal (MB) - 0.02 - 0.03 Palm Kernel Shell (PKS) - 0.04 - 0.05 Palm Mesocarp Fibre (PMF) - 0.06 Empty Fruit Brunch (EFB) - 0.07 - 0.08 25 100 200 300 400 500 600 700 800 900 Temperature (oC) Fig. 1. (a) TG curves and (b) DTG curves for Mukah Balingian coal and oil palm biomass materials at heating rate of 10 °C/min. Mukah Balingian coal results in one major peak in the high temperature range of 380–580 °C. This peak was attributed to the release of carbon containing volatile matter (Ismail et al., 2005b). As for biomass, its thermal degradation can be divided into three stages; moisture drying, main devolatilisation and continuous slight devolatilisation (Munir et al., 2009). The moisture drying region (temperature range of less than 150 °C) corresponds to the first peak in the DTG evolution profiles of the oil palm biomass as indicated in Fig. 1(b). The main devolatilisation regime of the three palm biomass samples showed different characteristics. As can be observed, the PMF starts to degrade first (at temperature of 180 °C) followed by EFB and PKS at temperature of 200 °C, and the weight loss starts to decrease promptly after that temperature. The three oil palm biomasses have less observable evolution profiles between temperature ranges of 350–500 °C, indicated by the non-zero value on DTG curve after the main peak of devolatilisation. There is no obvious weight loss observed beyond temperature of 550 °C for both PMF and EFB. Nevertheless, the PKS continues to decompose until beyond temperature of 750 °C with maximum decomposition at temperature of 704 °C. Oil palm biomass is of lignocellulosic material which consist mainly hemicellulose, cellulose and lignin. Apparently, the thermal decomposition of EFB showed the presence of only one major DTG evolution profile at maximum temperature (Tmax) of about 300 °C. Abdullah and Gerhauser (2008) reported the thermal evolution profile of EFB appeared at much higher temperature, i.e. 355 °C and was assigned to the thermal decomposition of cellulose (ca. 297–327 °C) component appeared in the samples (Várhegyi et al., 1997; Yang et al., 2007). As expected, the lower temperature shoulder appeared in DTG curve of palm mesocarp fibre (PMF) represents the decomposition of hemicellulose material and the higher temperature peak assigns to decomposition of cellulose material (Kastanaki et al., 2002). With regards to the thermal decomposition of PKS, the first and second DTG evolution profiles is expected to correspond to the devolatilisation of hemicellulose and cellulose materials, respectively (Kastanaki et al., 2002; Vamvuka et al., 2003). The non-observable peak exists in the three biomass samples was attributed by the decomposition of lignin which is known to decompose slowly over a wide range of temperature (137–667 °C) (Vamvuka et al., 2003). Nonetheless, the third peak which appeared to be completely isolated at temperature over 650 °C is not reported by other investigations on similar samples. The repetitiveness occurrence of this peak is further analysed via TG/MS (Mass Spectrometer) analysis. The DTG curve with relative intensities for PKS thermal decomposition is plotted as in Fig. 2. It appeared that, only relative intensities of mass/charge ratio (m/z) of 2, 12, 16, 44 and 45 showed significant peak at temperature range between 600 °C and 750 °C. The co-existence of both peak having m/z of 16 and 12 was assigned for the presence of methane, while the peak correspond to m/z of 2 showed that hydrogen being released at this temperature range, both as a result of charring process (Gómez et al., 2005). The mass spectrometry intensity for m/z of 12, 44 and 45 were assigned for the release of carbon dioxide during the course of pyrolysis of PKS. This numerous high-temperature mass losses with the release of CO2 are indicative of the decomposition of mineral carbonates contained in PKS. Moreover, the 4587 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 peaks present in the DTG curves of decomposition of the coal and biomass materials is denoted as TEP (thermal evolution profile), with TEP 1 and 2 indicate the evolution profiles due to decomposition of biomass components, while TEP 3 and 4 indicate the evolution profile for decomposition of coal and inorganic materials present in PKS, respectively. It was observed that the highest rate of decomposition was obtained for EFB, while the temperature corresponding to the maximum pyrolysis rate was lower for PMF, indicating its higher reactivity among the samples (Kastanaki et al., 2002). On the other hand, the thermal decomposition of Mukah Balingian (MB) coal starts at about 250 °C, temperature which is higher than that of biomass materials. The maximum pyrolysis rate was achieved at temperature of 432 °C at a mean reactivity of 6 times lower than that of biomass materials, thus as expected, this signify that MB coal is less reactive. The release of volatile matter in the MB coal ends in wide temperature interval in comparison to biomass, as indicated by the non-zero value of the DTG curves. In addition, its volatile matter amounted to 32 wt.% on dry basis, which is much lower with comparison to biomass samples with the highest amount of 72 wt.% released by PKS decomposition. Fig. 2. DTG curve with MS intensities for decomposition of palm kernel shell (PKS) at heating rate of 10 °C/min. results on mineral analysis using XRF reported in Table 2, proved that, the mineral content particularly calcium present in enormous amount as compared to other elements in EFB and PMF. The high amount of calcium carbonates present in PKS samples is due to the addition of this substance during the final separation process in the oil palm mill. Similar observation was reported elsewhere on thermal decomposition of biogran (Biagini et al., 2002), lignite (Vamvuka et al., 2003), and chicken litter (Whitely et al., 2006). The peak position and height in the DTG evolution profiles represent the reactivity of the samples i.e. the temperature corresponding to peak height is inversely proportional to reactivity while the peak height is directly proportional to the reactivity (Vamvuka et al., 2003). Table 3 shows the reactivity and its corresponding temperature at maximum rate of decomposition for coal and oil palm biomass materials at heating rate of 10 °C/min. The 3.2. Thermal decomposition of coal/PMF, coal/PKS and coal/EFB blends Fig. 3 represents the DTG curves for the thermal decomposition of MB coal/oil palm biomass for various blends at heating rate of 10 °C/min. Each of the oil palm biomass to coal blends shows a distinctive DTG evolution profiles. The coal/EFB blends (Fig. 3a) revealed two evolution profiles with the first peak at temperature (Tmax) of 300 °C, while the second peak at temperature 430 °C. Three thermal evolution profiles showed in the thermal decomposition of coal/PMF blend (Fig. 3b), with the first, second and third peaks at temperatures of 277 °C, 336 °C and 428 °C, respectively. As for coal/PKS blend (Fig. 3c), four thermal evolution profiles present during the pyrolysis reaction with the first to fourth peak temperature of 284 °C, 353 °C, 433 °C and 699 °C, respectively. The comparison between the single biomass and coal fuel with that of the coal/biomass blends for the three oil palm biomass samples were shown in Fig. 3. One may regards the first peak in coal/EFB blends corresponds to the EFB decomposition, while the Table 3 The maximum temperature and reactivity of Mukah Balingian coal, EFB, PKS, PMF, and their respective blends at heating rate of 10 °C/min. Biomass materials BLEND MB:BM TEP 1 Tmax EFB PMF PKS TEP 2 R 100:0 80:20 60:40 50:50 40:60 20:80 0:100 TEP 3 Tmax R 300 301 300 300 301 301 0.18 0.52 0.68 0.78 1.05 1.60 100:0 80:20 60:40 50:50 40:60 20:80 0:100 277 279 277 275 276 277 0.06 0.08 0.13 0.15 0.20 0.25 336 337 337 336 336 336 0.14 0.30 0.41 0.50 0.63 0.60 100:0 80:20 60:40 50:50 40:60 20:80 0:100 284 284 284 283 284 282 0.06 0.13 0.14 0.20 0.25 0.39 353 352 352 353 352 352 0.19 0.36 0.41 0.52 0.68 0.76 TEP 4 Rm VM 0.24 0.14 0.09 0.06 0.05 0.02 0.06 0.09 0.19 0.24 0.27 0.35 0.53 38.70 42.55 51.17 54.52 56.86 60.71 69.82 432 428 427 428 432 428 0.24 0.11 0.07 0.05 0.04 0.02 0.06 0.09 0.13 0.18 0.21 0.26 0.27 38.81 43.33 49.04 53.59 56.65 62.87 69.87 432 433 431 428 426 418 0.24 0.09 0.06 0.04 0.03 0.02 0.06 0.10 0.17 0.18 0.24 0.30 0.38 38.81 46.05 52.35 51.31 57.21 61.45 71.15 Tmax R 432 431 429 425 426 433 Tmax 679 691 695 695 699 704 R 0.04 0.07 0.08 0.10 0.14 0.18 TEP = thermal evolution profile; BM = biomass material; PMF = palm mesocarp fibre; PKS = palm kernel shell; EFB = empty fruit bunch; R: peak height (mg/min); Tmax = temperature of maximum weight loss (°C); VM = volatile matter released (wt.%). 4588 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Derivative weight (% / min) (a) 0.00 -0.025 MB MB(50): EFB(50) MB(80): EFB (20) MB(40): EFB(60) -0.05 Increasing wt% of biomass Increasing peak height (R) MB(60): EFB(40) MB(20): EFB(80) EFB -0.075 50 100 200 300 400 500 600 700 800 900 Temperature ( oC) Derivative weight (% / min) (b) 0.00 MB -0.025 MB (50): PMF (50) MB (80): PMF (20) MB (40): PMF (60) Increasing wt% of biomass Increasing peak height (R) MB (60): PMF (40) -0.05 MB (20): PMF (80) PMF 50 100 200 300 400 500 600 700 800 900 Temperature ( oC) 0.02 Derivative weight (% / min) (c) 0.00 MB (50): PKS (50) MB100 -0.02 MB (40): PKS (60) MB (80): PKS (20) -0.04 MB (20): PKS (80) Increasing wt% of biomass Increasing peak height (R) MB (60): PKS (40) PKS 100 50 100 200 300 400 500 600 700 800 900 Temperature ( oC) Fig. 3. DTG curves of (a) coal/EFB, (b) coal/PMF and (c) coal/PKS for various blends at a heating rate of 10 °C/min. second peak due to that of coal decomposition. Similarly, the first two peaks present in coal/PMF blends belongs to the PMF present in the blends while peak at higher temperature was as a results of the coal present in the blend. The same characteristic is observed in the coal/PKS blends with the fourth peak observed in the blends was attributed by the inorganic carbonates material in the biomass sample. Likewise, the observed thermal evolution profiles in the coal/biomass blends could be attributed to the decomposition of polymers of hemicellulose, cellulose and lignin present in biomass fuels (Jones et al., 2005; Kastanaki et al., 2002; Vamvuka et al., 2003). It was found that the DTG peak height (R) for the first evolution profile increased with increasing percentage of the biomass in the blends as depicted in Table 3. The same behaviour was observed with the second evolution profile in both coal/PKS and coal/PMF blends, and in the fourth evolution profile of coal/PKS blend. It appeared that no obvious shift in pyrolysis temperature in the blends with respect to that observed for the individual fuel. Furthermore, the increment in peak height appears to correlate with the mass ratios of biomass in the fuel blends. Nevertheless, it is interesting to note that fourth evolution profile for the co-pyrolysis of coal/PKS seem affected with the presence of coal in the blends. The evolution of this peak appeared at lower temperature than that of the raw PKS, which could be due to the catalytic effect as a result of the presence of considerable amount of inorganic carbonates in the sample, as indicated in the XRF analysis (Table 2). Yet the thermal evolution profile due to the release of volatile matter from coal decomposition in the blends showed a slight decreased in peak height with increasing in the percent of biomass in the blends which signify that the presence of oxygenated species released by the biomass pyrolysis does not affect the pyrolysis of coal (Biagini et al., 2006). The higher yield of volatiles, in biomass materials are attributed to stronger effects of temperature and heating rate in the depolymerisation reaction, which leads to rapid evolution of volatile content of the biomass materials. On contrary, the lower volatile release in coal is most likely due to the stronger bonding within the molecular structure of coal (Meesri and Moghtaderi, 2002). In an attempt to investigate possible synergy in the coal/biomass blends, the amount of volatile matter released from the 4589 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 ð1Þ where xMB and xOPW is the percentage of MB coal and oil palm wastes in the blend, respectively, and WMB and WOPW is the normalised weight loss of MB coal and oil palm wastes, respectively, under the same operative conditions. Calculated and experimental weight loss curves were compared as shown in Fig. 5. The slight differences shown were within the experimental error range. The results agree reasonably well with those of similar studies done by other researchers who co-pyrolysed coal with cotton residue, olive kernel, forest residue, rice husk, sugarcane bagasse, pine saw dust and pine (Biagini et al., 2002; Ismail et al., 2005a; Jones et al., 2005; Vamvuka et al., 2003; Vuthaluru, 2004). 3.3. Effect of heating rates on reactivity of coal, biomass materials and their blends (a) 90 MB coal Normalised weight loss,% W blend ¼ xMB W MB þ xOPW W OPW 100 80 70 calc. 50: 50 MB/EFB 60 50 40 exp. 50: 50 MB/EFB 30 EFB 20 10 0 0 100 200 300 400 500 600 700 800 900 1000 Temperature, o C 100 90 (b) MB coal 80 Normalised weight loss,% co-pyrolysis reaction of coal and palm biomass were plotted against the proportion of biomass in the blends as shown in Fig. 4. The lack of synergy between the two fuels during co-pyrolysis is indicated by the linear relationship between the volatile matter release and the percentage of biomass added to the mixture (Jones et al., 2005; Vuthaluru, 2004) and this condition was observed on all the blends. It was also possible to verify this behaviour using the weighted average of normalised weight loss of the blend, as shown in Eq. (1) 70 60 calc. 50: 50 MB/PMF 50 40 exp.50: 50 MB/PMF 30 PMF 20 10 0 0 100 200 300 400 500 600 700 800 900 1000 Temperature,o C 100 90 (c) MB coal 80 Normalised weight loss, % Table 4 shows the Tmax, and peak height (R) at various heating rates (i.e. 10, 20, 40, 60 °C/min) for oil palm biomass materials (PMF, PKS and EFB), Mukah Balingian coal and coal/biomass blends at 50 wt.%. Evidently, Tmax and R for the evolution profiles of thermal decomposition of palm biomass materials, coal and their blends, increase with increasing heating rates during the pyrolysis process, indicating an increase in reactivity. One may observe this behaviour from the thermograms of EFB, PKS and PMF, and coal, where the peak was systematically shifted to higher temperature regime as the heating rate increases. Fig. 6 describes the behaviour of thermograms of PKS at various heating rates. Similar trends were observed with PMF, Mukah Balingian coal and 50:50 coal/ palm biomass blends. The reason for these temperature shifts is that less heat is required for the cracking of the solid fuel particles into products. This process is reached later at higher temperatures since the heat transfer is not as effective and efficient as they were at lower heating rates. At lower heating rates, the heating of biomass particles occurred more gradually leading to an improved and more effective heat transfer to the inner portions and among the parti- 70 60 calc. 50: 50 MB/PKS 50 40 exp. 50: 50 MB/PKS 30 PKS 20 10 0 0 100 200 300 400 500 600 700 800 900 1000 Temperature, o C Volatile matter released (wt%) Fig. 5. Comparison of calculated and experimental TG curves for (a) MB/EFB blend, (b) MB/PMF blend and (c) MB/PKS blend at heating rate 10 °C/min. MB/PKS MB/PMF MB/EFB Linear (MB/PMF) Linear (MB/PKS) Linear (MB/EFB) Percentage of biomass in coal/biomass blend (wt%) Fig. 4. Volatile matter released from coal/biomass blends at different percentage of biomass in coal blends. cles (Biagini et al., 2006). Therefore cracking process was more effective which caused more weight loss in the form of volatiles components. As the heating rate is decreased, the residue at the end of pyrolysis reactions is also decreased. The Tmax, which is the point at which maximum weight loss takes place, shifts to higher values with an increase in the heating rates. This shifts have been reported by researcher on other biomass materials (Biagini et al., 2008; Luangkiattikhun et al., 2008; Vamvuka et al., 2003). In this study, it was found that the amount of percent volatile matter released (VM) differs between the biomass and coal, but not much difference among the biomass fuels and this shows that heating rates have less impact on the release of volatile matter, thus the char yield. The same observation was reported by Luangkiattikhun et al. (2008) on pyrolysis of PKS and PMF at various heating rates. 4590 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Table 4 Peak temperature (Tmax) and peak height at heating rates of 10, 20, 40 and 60 °C/min for biomass materials, coal and coal/biomass (50:50) blends. FEED HR TEP 1 EFB 10 20 40 60 PMF 10 20 40 60 277 288 298 304 PKS 10 20 40 60 282 295 306 313 MB 10 20 40 60 MB:EFB (50:50) 10 20 40 60 MB:PMF (50:50) 10 20 40 60 277 289 298 305 MB:PKS (50:50) 10 20 40 60 284 295 308 316 TEP 2 R Tmax TEP 3 TEP 4 Tmax R Tmax VM Tmax R 301 311 320 325 1.60 2.89 6.23 9.04 69.8 63.4 64.5 64.0 0.25 0.64 1.17 2.00 336 348 358 377 0.6 1.09 1.81 2.4 69.9 69.9 66.6 67.3 0.39 0.70 1.40 2.11 352 363 375 382 0.76 1.26 2.12 2.87 704 724 745 754 R 0.18 0.36 0.62 0.92 71.2 69.0 65.5 66.8 432 442 453 460 0.24 0.62 1.04 1.63 38.7 38.3 37.7 38.5 300 311 320 326 0.68 1.33 2.22 3.89 425.5 437.4 446.1 453.4 0.06 0.14 0.27 0.31 54.5 51.2 50.5 49.9 0.13 0.25 0.43 0.84 337 348 358 361 0.41 0.75 1.23 1.75 428 437 450 455 0.05 0.1 0.19 0.28 53.6 52.3 51.7 52.4 0.14 0.28 0.62 1.10 352 365 377 385 0.41 0.72 1.31 2.00 428 440 451 456 0.04 0.08 0.17 0.21 695 711 736 751 0.08 0.18 0.36 0.60 51.3 49.8 52.9 54.8 TEP = thermal evolution profile; HR = heating rate; BM = biomass material; PMF = palm mesocarp fibre; PKS = palm kernel shell; EFB = empty fruit bunch; R: peak height (mg/ min); Tmax = temperature of maximum weight loss (°C); VM = volatile matter released (wt.%). mogravimetric analysis at constant heating rate, Hr = dT/dt, thus Eq. (2) may be written as Eq. (3); 5   da A Ea exp ¼ f ðaÞ dT Hr RT Derivative Weight (%/min) 0 -5 -10 At the peak temperature (Tmax), where the maximum rate of decomposition is reached, the second derivative of Eq. (2) is equal to zero and is given by Eq. (4) o 10 C/min 20oC/min -15 -20 ð3Þ 2 40oC/min d a 60oC/min dt -25 2 ¼ " Ea H r RT 2max ! #  Ea da 0 þ A exp ¼0 f ðaÞ RT max dt  ð4Þ where f0 (a) is the derivative form of f(a). Thus after simple rearrangement of Eq. (4), the equation becomes Eq. (5) -30 Hr -35 25 100 200 300 400 500 600 700 800 900 Temperature, oC ¼     AR Ea exp f 0 ðaÞ Ea RT max ð5Þ It is transformed into linear function as in Eq. (6) by taking natural log on both sides of Eq. (5) Fig. 6. DTG curves of PKS at heating rate of 10, 20, 40 and 60 °C/min. 3.4. Kinetic analysis ln The activation energy and pre-exponential factor of coal and biomass pyrolysis were determined by differential method (Bamford and Tipper, 1980). The fundamental rate equation used in all kinetic studies is expressed as Eq. (2);   da Ea ¼ kðTÞf ðaÞ ¼ A exp f ðaÞ dt RT T 2max ! ð2Þ where k is the rate constant and f(a) is the reaction model, a function depending on the actual mechanism. For non-isothermal ther- Hr T 2max ! ¼      Ea 1 AR nð1  aÞn1 þ ln T max Ea R ð6Þ In this study, it is assumed that solid fuel pyrolysis is a first-order reaction (Biagini et al., 2002). Although it is not always the case, but the simplicity and the number of parameters of the first-order model allow the direct comparison of different fuels to provide preliminary parameters for further and more accurate modelling. Thus, Eq. (6) may be written as Eq. (7) ln Hr T 2max ! ¼      Ea 1 AR þ ln T max Ea R ð7Þ 4591 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Table 5 The activation energy (Ea), pre-exponential factor (A) and calorific value of coal, biomass materials and their respective blends. Feed Blend MB: EFB 100:0 80:20 60:40 50:50 40:60 20:80 0:100 Ea (kJ/mol) A (s1) R2 (%) TEP 1 MB: PMF MB: PKS Ea (kJ/mol) A (s1) R2 (%) TEP 2 A (s1) R2 (%) 272.88 266.61 263.60 262.95 243.24 239.13 1.74E + 18 6.54E + 17 4.39E + 17 4.78E + 17 1.45E + 16 4.86E + 15 99.77 99.68 99.67 99.62 99.98 99.87 24.02 23.09 21.53 20.69 19.88 18.19 16.58 272.88 258.71 271.37 266.04 263.66 262.84 1.74E + 18 1.87E + 17 1.87E + 18 7.02E + 17 4.23E + 17 4.05E + 17 99.77 97.83 99.43 99.72 99.24 99.92 24.02 23.53 22.08 21.38 20.49 19.22 17.80 272.88 254.93 264.34 261.69 233.97 223.72 1.74E + 18 7.86E + 16 4.61E + 17 3.07E + 17 2.83E + 15 5.56E + 14 99.77 99.99 99.92 99.26 98.29 98.28 Ea (kJ/mol) TEP3 181.55 189.92 194.11 207.12 226.85 209.68 4.01E + 14 2.20E + 15 5.47E + 15 9.40E + 16 5.80E + 18 1.45E + 17 99.82 99.86 99.89 99.73 99.81 99.93 100:0 80:20 60:40 50:50 40:60 20:80 0:100 157.71 161.48 162.55 163.83 165.64 166.24 9.49E + 12 2.20E + 13 2.82E + 13 4.19E + 13 6.13E + 13 6.55E + 13 89.55 99.42 98.78 98.53 99.87 99.80 198.19 202.59 206.97 202.62 205.55 216.10 1.01E + 15 2.30E + 15 5.94E + 15 2.60E + 15 4.56E + 15 3.54E + 16 99.16 98.30 98.45 98.78 99.23 98.78 100:0 80:20 60:40 50:50 40:60 20:80 0:100 138.97 137.84 142.72 144.66 149.59 150.13 9.28E + 10 7.47E + 10 2.27E + 11 3.54E + 11 1.09E + 12 1.16E + 12 98.56 99.52 99.91 99.79 99.91 99.79 174.35 176.43 176.43 187.64 186.35 192.12 3.01E + 12 4.93E + 12 4.93E + 12 4.46E + 13 3.53E + 13 1.12E + 14 99.09 99.93 99.98 99.92 99.97 99.98 Ea (kJ/mol) A (s1) R2 (%) LHV (MJ/kg) TEP 4 234.74 251.29 245.79 255.61 260.58 277.11 3.51E + 10 2.17E + 11 1.01E + 11 3.18E + 11 5.39E + 11 3.72E + 12 97.87 99.97 99.65 98.45 99.64 99.66 24.02 23.23 21.62 20.74 19.80 18.12 16.72 TEP = thermal evolution profile; PMF = palm mesocarp fibre; PKS = palm kernel shell; EFB = empty fruit bunch; R2 = correlation coefficient; Ea = activation energy (kJ/mol); A = pre-exponential factor (s1), LHV = low heating value (MJ/kg). where a = fraction decomposition, A = pre-exponential factor, Ea = activation energy, R = gas constant, n = reaction order, T = temperature at peak, Hr = heating rate. Eq. (6) is the common ‘‘modelfree” kinetics equation that is first developed by Kissinger (Kissinger, 1956). In the Kissinger method, lnðHr =T 2max Þ is plotted against (1/ Tmax) for a series of experiments performed using TGA at different heating rates with peak temperature obtained from the DTG curve. In this experiment, activation energy, Ea, pre-exponential factor, A and correlation coefficient (R2) obtained from the pyrolysis and co-pyrolysis of PMF, PKS, EFB, coal and their respective blends are presented in Table 5. In general, the value of activation energies and pre-exponential factors for the first thermal evolution profile (TEP 1), increases with the addition of biomass in the coal blends. Similar trend was observed in the second thermal evolution profile (TEP 2) for coal/ PMF and coal/PKS blends with an exception for coal/EFB mixtures. In comparison, the activation energies due to the decomposition of cellulose (TEP 2) present in biomass showed the highest value (174–227 kJ/mol), than that due to the decomposition of hemicellulose (TEP 1) (ca. 140–166 kJ/mol). These values fall within the range of that reported elsewhere on olive kernel (Kastanaki et al., 2002), rice husk (Ismail et al., 2005a) and corn cob (Fisher et al., 2002). Evidently, the activation energy of the co-pyrolysis reaction due to the evolution of volatile matter in coal (TEP 3) is minimum at 80 wt.% of biomass in the coal feed stream for both coal/EFB and coal/PKS. As of coal/PMF blend, the minimum value of activation energy is reported for 20 wt.% of PMF in the coal blend. The activation energy of the coal blends rises to a peak of 266.6 kJ/mol, 271.4 kJ/mol and 264.3 kJ/mol at 20 wt.% biomass for coal/EFB mixture and 40 wt.% biomass content for both coal/PKS and coal/ PMF mixtures, respectively. The present trend in the activation energies and pre-exponential factors for the coal/biomass blends seems to be in good agreement with the ones obtained by other researchers (Ismail et al., 2005a; Vuthaluru, 2004). As the energy barrier, activation energy provides the information of critical energy needed to start a reaction. Thus, to ensure lower activation en- ergy and so lower temperatures required for promotion of forward pyrolysis reaction, the mixture with lowest activation energy is recommended. However, the heating value of the fuel blend should be considered to be over 20 MJ/kg as to ensure auto-thermal combustion (Biagini et al., 2002). Based on this limitation, it can be concluded that, the optimum blends based on the lowest activation energy on coal thermal evolution profile and the acceptable heating value limit, are 50: 50 for coal/EFB and 80:20 for both coal/ PMF and coal/PKS. Comparing activation energy data has been a common method of analysis for the suitability of various biomass types to be converted into renewable fuel source and has been used by various researchers on coal/wood wastes (Vuthaluru, 2004) and rice husk (Mansaray and Ghaly, 1999). Apparently, the kinetic parameters reported for the forth thermal evolution profiles (TEP 4) in pyrolysis of coal/PKS were slightly higher values as compared to the values found in the literature (Liu et al., 2003; Zhao et al., 2001). These differences could possibly due to the use of different model that describe the decomposition behaviour of calcium carbonates (Ioannou et al., 2009; Maciejewski, 2000). However, further studies have to be carried out in order to understand this behaviour. 4. Conclusions Thermal behaviour of Malaysian sub-bituminous coal and three oil palm biomass materials have been investigated using thermogravimetric analysis. Evidently, the three palm biomass wastes and coal showed distinct behaviour upon pyrolysis. As expected the pyrolysis of the PMF, PKS and EFB occurred at lower temperature in comparison to coal sample. The mean reactivity is in the order of EFB > PKS > PMF > MB coal. Apparently, the coal/oil palm biomass blends appear to undergo an independent thermal degradation without any synergistic interaction as such its behaviour during co-pyrolysis can be estimated from those of parent fuels. Nevertheless, the presence of biomass in the coal blend represents carbon neutrality for further development of co-firing system in Malaysia. 4592 S.S. Idris et al. / Bioresource Technology 101 (2010) 4584–4592 Acknowledgements The authors would like to thank the Research Management Institute (RMI), University Technology MARA, Faculty of Chemical Engineering and Biomass and Fossil Fuel Research Team (UiTM Perlis) for their support. References Abdullah, N., Gerhauser, H., 2008. Bio-oil derived from empty fruit bunches. Fuel 87, 2606–2613. Bamford, C.H., Tipper, C.F.H. (Eds.), 1980. Decomposition reactions of solids. In: Green, N. (Ed.), Comprehensive Chemical Kinetics, vol. 22. Elsevier, New York, pp. 115–246. Bernama, 2008. Malaysia to Develop Renewable Energy Action Plan. (Retrieved 9 January 2009) <http://www.ktak.gov.my/template03.asp?tt=news&newsID=420>. Biagini, E., Barontini, F., Tognotti, L., 2006. Devolatilization of biomass fuels and biomass components studied by tg/ftir technique. Industrial Engineering and Chemical Research 13, 4486–4493. Biagini, E., Fantei, A., Tognotti, L., 2008. Effect of the heating rate on the devolatilization of biomass residues. Thermochimica Acta 1–2, 55–63. Biagini, E., Lippi, F., Petarca, L., Tognotti, L., 2002. Devolatilization rate of biomasses and coal-biomass blends: an experimental investigation. Fuel 81, 1041–1050. Caballero, J.A., Conesa, J.A., Font, R., Marcilla, A., 1997a. Pyrolysis kinetics of almond shells and olive stones considering their organic fractions. Journal of Analytical and Applied Pyrolysis 2, 159–175. Caballero, J.A., Marcilla, A., Conesa, J.A., 1997b. Thermogravimetric analysis of olive stones with sulphuric acid treatment. Journal of Analytical and Applied Pyrolysis 1, 75–88. Chuah, T.G., Azlina, A.G.K.W., Robiah, Y., Omar, R., 2006. Biomass as the renewable energy sources in Malaysia: an overview. International Journal of Green Energy 3, 323–346. Fisher, T., Hajaligol, M., Waymack, B., Kellogg, D., 2002. Pyrolysis behaviour and kinetics of biomass derived materials. Journal of Analytical and Applied Pyrolysis 2, 331–349. Gómez, C.J., Velo, E., Puigjaner, L., 2005. Comparative thermogravimetry/mass spectrometry study of woody residuals and an herbaceous biomass crop using PCA techniques. In: Paper Presented at the American Institute of Chemical Engineer (AIChE) Annual Meeting, Ohio, USA. Ioannou, Z., Zoumpoulakis, L., Halikia, I., Teloniati, T., 2009. Overall kinetic study of non-isothermal decomposition of calcium carbonate. Mineral Processing and Extractive Metallurgy 2, 98–104. Ismail, K., Zakaria, Z., Ishak, M.A.M., 2005a. Thermal behaviour study of Mukah Balingian coal and biomass blend during pyrolysis via thermogravimetric analysis. In: Paper Presented at the 22nd International Pittsburgh Coal Conference, Pittsburgh USA. Ismail, K., Zakaria, Z., Ishak, M.A.M., 2005b. Thermal behaviour study of Mukah Balingian coal and rice husk blends during pyrolysis via thermogravimetric analysis. In: Paper Presented at the Brunei International Conference on Engineering and Technology, BICET 2005, Brunei. Jones, J.M., Kubacki, M., Kubica, K., Ross, A.B., Williams, A., 2005. Devolatilisation characteristics of coal and biomass blends. Journal of Analytical and Applied Pyrolysis 75, 502–511. Kalam, A., Shahari, B., Khalid, Y., Shaw, W., Jabatan, V., 2004. Oil palm fruit bunch fiber composite. In: Paper Presented at the International SAMPE Technical Conference. Kastanaki, E., Vamvuka, D., Grammelis, P., Kakaras, E., 2002. Thermogravimetric studies of the behavior of lignite–biomass blends during devolatilization. Fuel Processing Technology 77–78, 159–166. 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 11, 5692–5701. Kissinger, H.E., 1956. Variation of peak temperature with heating rate in differential thermal analysis. Journal of Research of the National Bureau of Standards 217, 221. Liu, R., Chen, J., Guo, F., Yun, J., Shen, Z., 2003. Kinetics and mechanism of decomposition of nano-sized calcium carbonate under non-isothermal condition. Chinese Journal Chemical Engineering 3, 302–306. Luangkiattikhun, P., Tangsathitkulchai, C., Tangsathitkulchai, M., 2008. Nonisothermal thermogravimetric analysis of oil-palm solid wastes. Bioresource Technology 5, 986–997. Maciejewski, M., 2000. Computational aspects of kinetic analysis: part b: the ictac kinetics project – the decomposition kinetics of calcium carbonate revisited, or some tips on survival in the kinetic minefield. Thermochimica Acta 355, 145– 154. Mansaray, K.G., Ghaly, A.E., 1999. Kinetics of the thermal degradation of rice husks in nitrogen atmosphere. Energy Sources, Part A: Recovery, Utilization, and Environment Effects 21, 773–784. Meesri, C., Moghtaderi, B., 2002. Lack of synergetic effects in the pyrolytic characteristics of woody biomass/coal blends under low and high heating rate regimes. Biomass and Bioenergy 1, 55–66. Moghtaderi, B., Meesri, C., Wall, T.F., 2004. Pyrolytic characteristics of blended coal and woody biomass. Fuel 6, 745–750. Munir, S., Daood, S.S., Nimmo, W., Cunliffe, A.M., Gibbs, B.M., 2009. Thermal analysis and devolatilization kinetics of cotton stalk, sugar cane bagasse and shea meal under nitrogen and air atmospheres. Bioresource Technology 3, 1413–1418. Pan, Y.G., Velo, E., Puigjaner, L., 1996. Pyrolysis of blends of biomass with poor coals. Fuel 75, 412–418. Rahman Mohamed, A., Lee, K.T., 2006. Energy for sustainable development in Malaysia: energy policy and alternative energy. Energy Policy 15, 2388– 2397. Vamvuka, D., Kakaras, E., Kastanaki, E., Grammelis, P., 2003. Pyrolysis characteristics and kinetics of biomass residuals mixtures with lignite. Fuel 82, 1949– 1960. Várhegyi, G., Antal, M.J., Jakab, E., Szabó, P., 1997. Kinetic modeling of biomass pyrolysis. Journal of Analytical and Applied Pyrolysis 1, 73–87. Vuthaluru, H.B., 2004. Investigations into the pyrolytic behaviour of coal/biomass blends using thermogravimetric analysis. Bioresource Technology 2, 187–195. Whitely, N., Ozao, R., Cao, Y., Pan, W.-P., 2006. Multi-utilization of chicken litter as a biomass source. Part II. Pyrolysis. Energy and Fuels 6, 2666–2671. Yang, H., Yan, R., Chen, H., Ho Lee, D., Zheng, C., 2007. Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel 86, 1781–1788. Yang, H., Yan, R., Chin, T., Liang, D.T., Chen, H., Zheng, C., 2004. Thermogravimetric analysis-Fourier transforms infrared analysis of palm oil waste pyrolysis. Energy and Fuels 6, 1814–1821. Yusoff, S., 2006. Renewable energy from palm oil – innovation on effective utilization of waste. Journal of Cleaner Production 14, 87–93. Zhao, Y.T., Sun, T.S., Sun, B., 2001. The thermal decomposition of calcium carbonate. Chinese Chemical Letters 8, 745–746.