- Email: [email protected]
biomass blends: A Turkish case study
Available online at www.sciencedirect.com
Waste Management 28 (2008) 2077–2084 www.elsevier.com/locate/wasman
Effect of co-combustion on the burnout of lignite/biomass blends: A Turkish case study H. Haykiri-Acma, S. Yaman
*
Istanbul Technical University, Chemical and Metallurgical Engineering Faculty, Chemical Engineering Department, 34469 Maslak, Istanbul, Turkey Accepted 22 August 2007 Available online 26 October 2007
Abstract Co-combustion of Turkish Elbistan lignite and woody shells of hazelnut was performed in a TGA up to 1173 K with a heating rate of 20 K/min. SEM images of each fuel revealed the differences in their physical appearances. Hazelnut shell was blended with lignite in the range of 2–20 wt% to observe the co-combustion properties. Maximum burning rates (Rmax), temperatures of the maximum burning rates (TR-max), and the final burnout values of the parent samples and the blends were compared. The results were interpreted considering lignite properties and the major biomass ingredients such as cellulosics, hemicellulosics, and lignin. Deviations between the theoretical and experimental burnout values were evaluated at various temperatures. Burnout characteristics of the blends up to 10 wt% were concluded to have a synergistic effect so the addition of hazelnut shell up to 8 wt% provided higher burnouts than the expected theoretical ones, whereas addition of as much as 10 wt% led to a decrease in the burnout. However, the additive effects were more favorable for the blend having a biomass content of 20 wt%. Apparent activation energy, Rmax, and TR-max, were found to follow the additive behavior for the blend samples. Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction Co-combustion is a common practice and an effective method for waste disposal, not only can significant energy recovery be obtained from waste materials but also a marked reduction in the volume of the wastes is realized. There are several reasons to blend waste materials with coal or with other types of fuels prior to burning. In particular, the ash deposition and fouling problems on hot surfaces, which are commonly encountered in combustion of biomass, can be diminished or eliminated by means of burning of biomass/coal blends. On the other hand, the ash fraction of various biomass species is usually rich in alkaline and alkaline earth elements, providing an important potential to capture the SO2 being released from the combustion of high sulfur coals (Diaz-Somoano et al., 2006; Martin et al., 2006).
*
Corresponding author. Tel.: +90 212 2856873; fax: +90 212 2852925. E-mail address: [email protected] (S. Yaman).
0956-053X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2007.08.028
Since fresh biomass fixes solar radiation and CO2 in the form of lignocellulosics during photosynthesis, the emitted CO2 from combustion of these lignocellulosic materials is regarded not to have any net contribution to the CO2 accumulation in atmosphere. Hence, compared to combustion of fossil fuels alone, usage of biomass-blended fuels is less harmful in terms of greenhouse effect (Klass, 1998). Besides, a fuel is accepted as non-corrosive if its S/Cl molar ratio is greater than 4, whereas fuels with S/Cl molar ratios that are lower than 2 lead to serious corrosion in materials (Ferrer et al., 2005). This ratio can be adjusted by coexisting biomass and coal in the combustion medium. Co-combustion of municipal solid wastes (MSW) and agricultural residues offers the possibility of reducing the need for landfill sites owing to the significant reductions in the amounts of the waste materials, and improvements in the ash properties of the blended fuels compared to that of the parent waste materials (Steenari and Lindqvist, 1999). In the case of refuse derived fuel (RDF) combustion, polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo-p-furans (PCDFs) emissions are formed,
2078
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
causing serious environmental and health concerns. However, co-firing of RDF with coal is reported to inhibit the formation of these dangerous emissions considerably (Ferrer et al., 2005). Also, Raghunathan and Gullett (1996) showed that SO2 injection was evidently capable of preventing the formation of PCDD/F. Likewise, the presence of coal during co-firing also provides some reduction in the emissions of PCDD/Fs resulting from the combustion of PVC (Sanchez-Hervas et al., 2005). Combustion of blended straw or waste wood with coal was reported to have reasonably low environmental effects (Hartmann and Kaltschmitt, 1999). On the other hand, characteristics of ashes released from the co-combustion of meat and bone meal (MBM) and coal was seen as appropriate to landfill without environmental risk (Gulyurtlu et al., 2005). The composition of ash from biomass materials is generally different from ash from coal (van Krevelen, 1993). Alkalinity is usually higher in biomass ashes. The typical ratio of base/acid is calculated as follows: Base Fe2 O3 þ CaO þ MgO þ Na2 O þ K2 O þ P2 O5 ¼ Acid SiO2 þ Al2 O3 þ TiO2
ð1Þ
This ratio for biomass often exceeds 1.0 and may be beyond 2.0, whereas it is usually lower than 1.0 in for coal. The base/acid ratio is an indication of fuel performance in association with problematic deposit formations (Tillman, 2000; Pronobis, 2006). Potassium is the key element in most biomass species, leading to the formation of the undesired depositions and fouling on the heat exchanger surfaces. However, the sulfur content of the fuel that biomass is blended with may react with potassium to form less harmful products. Besides, potassium in a gaseous form may react with alumina-silicates coming from the coal in the blend (Amand et al., 2006). Combustion of low rank coals with biomass in power stations is reported to reduce the emissions of greenhouse gases, NOx, and SOx (Martin et al., 2006). It is a common practice to fire various organic wastes with coal in order to reduce the volume of dangerous waste, heat recovery, and removal of toxic organics (Elled et al., 2006). In various blends with straw, peat behaved as a cleaning agent with respect to deposition rate during co-firing. This was attributed to the alumina-silicate rich character of peat ash (Theis et al., 2006). Hazelnut production is of great importance in the Turkish agricultural sector so that Turkey generates about 80% of the world’s hazelnuts (Topuz et al., 2005). Consequently, several 100,000 tonnes of hazelnut wastes are renewably formed every year. The high calorific value of hazelnut shells makes it a potential energy source. This fuel is generally used in domestic heating or in small-scale combustion systems near to the cultivation areas. Ash from hazelnut shell contains a K2O content of 30.4 wt% (db), which is almost 50% of the sum of the basic components, and this leads to a base/acid ratio of 2.03 which is somewhat high
compared to coal (Demirbas, 2002). Such a high concentration of K2O is the main drawback encountered during the combustion of this potential fuel in commercial combustors. On the other hand, Elbistan lignite is the most important national primary energy resource in Turkey with a reserve of 3.4 billion metric tonnes. However, it has quite a low calorific value, and high sulfur and high ash contents. Therefore, effective measures must be taken in order to reduce the environmental risks resulting from the combustion of Elbistan lignite. Nevertheless, the K2O content of Elbistan lignite is very low, and it is less than 1 wt% even in its fly ash (Ural, 2005). Calcite is the key mineral species in Elbistan lignite; 27.4 wt% of dry coal is calcite. Other important minerals are quartz (4.8%), illite (2.6%), pyrite (1.6%), kaolinite (0.5%), and aragonite (0.5%). Calcium consists of 11.1 wt% of dry Elbistan lignite (Karayigit et al., 2000). Phosphor is regarded to inhibit the retention of sulphur by calcium compounds (Elled et al., 2006). In this context, since the phosphor content of hazelnut shell is not as high as in RDF or MBM, this effect is not expected to occur during co-combustion of hazelnut shell and high sulphur coals. For the outlined reasons, Elbistan lignite and hazelnut shell seem to have important potential to compensate the disadvantages coming from each other. Biomass may contribute to the calorific value of the low rank-low quality coal, and lignite may behave as a cleaning agent in terms of fouling or depositions on convective heating surfaces resulting from biomass. However, apparent differences in the chemical and physical properties of both fuels may cause some inconveniences in the combustion systems due to the different combustion characteristics of each fuel. Furthermore, burnout characteristics of their blends must be taken into account as another challenge. Hence, the aim of this paper is to compare the combustion properties of each fuel under the same slow combustion conditions, and to investigate the co-combustion characteristics and the burnouts of their blends. 2. Experimental In the experiments, woody shells of Turkish hazelnuts were used as the biomass material. Elbistan lignite was selected as the coal sample. Both samples were milled and sieved to a particle size of <0.250 mm. The proximate analysis and the gross calorific value measurements of the samples were conducted according to ASTM standards. Ultimate analyses of the samples were carried out using an EuroEA3000 model elemental analyzer. Major organic components in hazelnut shell were determined by analytical manner. For this purpose, extractives were determined applying the benzene–ethyl alcohol extraction procedure according to ASTM D1105. The bulk remaining after this procedure was treated with the mixtures
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
of NaClO2 and acetic acid to isolate the holocellulose (cellulosics + hemicellulosics) content. The lignin content was determined by the van Soest method (1963); a-cellulose content was determined according to TAPPI T203 standard (1988). Scanning electron microscopy (SEM) tests were performed in a Joelä Model JSM-T330 operated at 25 kV and linked with an energy dispersive (EDS) attachment to study the physical appearances of the fuel samples. Lignite and hazelnut shell were mixed to obtain various blends having hazelnut shell in the ratios of 2, 4, 6, 8, 10, and 20 wt%. Slow combustion of both parent fuels and their blends were performed in a Shimadzu TG 41 Thermogravimetric Analyzer (TGA) with a cylindrical alumina crucible. In these experiments, 40 mg of sample was heated from ambient to 1173 K by a constant heating rate of 20 K/min under dry air flow of 40 mL/min. Then, the sample was held at 1173 K until no further mass loss was detected. Derivative thermogravimetric analysis (DTG) profiles, showing the burning rates of the samples, were derived from the TGA data; and the Coats–Redfern method (1964) was applied to calculate the apparent activation energy values during the process. All experiments were repeated, and the mean values were used provided that the deviations were within 5%. 3. Results and discussion 3.1. Differences between the properties of the samples Table 1 presents the results of the proximate and ultimate analyses, calorific values of both samples, and major biomass components of hazelnut shell. Elbistan lignite is a low grade coal. Low fixed carbon and high ash contents are its typical characteristics. Besides, a sulfur content of as much as 3.6 wt% (daf) is seriously high,
2079
and effective measures must be taken in order to get rid of the problems resulting from SO2 emissions. On the other hand, its calorific value is considerably low. Having high sulfur, high ash, low fixed carbon, and low calorific value makes Elbistan lignite extremely problematic. Unlike Elbistan lignite, hazelnut shell has a relatively higher calorific value, lower ash content, and negligible sulfur that could not be detected by the analyzer used. More than half of the organic part of hazelnut shell is comprised from lignin. This is consistent with the woody structure of this sample; a-cellulose is the richest constituent in the holocellulose. However, the oxygen content of hazelnut was higher than that of Elbistan lignite. This indicates that this biomass material would have higher thermal reactivity than lignite. SEM images of the samples having particle sizes of <0.250 mm are illustrated in Fig. 1. According to these images, it can be said that there are fundamental differences in the physical appearances of both samples. These differences can be summarized as follows: Lignite is very rich in small particles in such a way that some particles with the sizes of as low as several microns can be easily noticed. This implies the fragile structure and the tendency toward dust formation of Elbistan lignite during milling. In fact, this is an important problem in the mining of this lignite. Whereas, small particles are relatively less in the case of hazelnut shell, owing to its fibrous structure. Lignite particles are seen to have spherical shape in general, and the borders of the particles are relatively regular. Although, hazelnut particles have a distinctly amorphous shape, they may be assumed to be rather cylindrical. In addition to the shape, the surface areas of the different size particles show fundamental differences for each sample. Some bright parts are available on the surfaces of the lignite particles, which may be attributed to the reflection of the mineral phases found in the lignite.
Table 1 Analyses results of hazelnut shell and Elbistan lignite Analysis
Hazelnut shell
Lignite
Moisture (%)
11.8
45.0
Volatile matter (%, db) Fixed carbon (%, db) Ash (%, db)
87.3 8.4 4.3
50.5 17.2 32.3
C (%, daf) H (%, daf) N (%, daf) S (%, daf) Oa (%, daf)
46.3 5.8 0.4 <0.05 47.5
53.2 5.5 1.7 3.6 36.0
Gross calorific value (kcal/kg, db) Extractives (%, db) Holocellulose (%, db) Lignin (%, db) a-Cellulose (%, db) a
Calculated by difference.
4358 6.2 38.6 51.5 22.9
2400 n/a n/a n/a n/a
3.2. Combustion properties of the parent samples TGA and DTG burning profiles of the samples are given in Figs. 2 and 3, respectively. Drying, thermal decomposition, volatile release, combustion of volatiles, char gasification, and char oxidation stages caused mass losses from the solid matrices of the samples. Before the initiation of the burning, some increase in mass could be observed due to the chemisorption of oxygen. Folgueras et al. (2003) also reported a similar increase in the mass of the sample during such an experiment. Maximum burning rates (Rmax), temperatures of the maximum burning rates (TR-max), and final burnout values were obtained from these profiles, and considered to compare the thermal behaviors of the samples under investigated combustion conditions.
2080
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Fig. 1. SEM images of the Elbistan lignite and hazelnut shell samples.
dm/dt (mg/min)
m (mg)
30
20
10
0 200
6
Lignite Biomass
40
400
600
800
1000
Temperature (K) Fig. 2. TGA curves of the samples.
1200
Lignite Biomass
4
2
0 200
400
600
800
1000
Temperature (K) Fig. 3. DTG profiles of the samples.
1200
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Fig. 2 shows that mass losses occurred from the samples upon heating. In spite of the relatively low temperatures, important structural decompositions occurred, especially in the hazelnut shell just after moisture release, and consequently marked devolatilization yields were detected. Holocellulosics in the hazelnut shell is responsible for these mass losses due to their high reactivity and weak structures (Haykiri-Acma et al., 2006). After that, cellulosics contributed to the formation of volatiles. Ignition of combustible volatiles augmented the mass losses from the sample. Thus, the rate of mass loss on Fig. 3 reached 5.50 mg/min at 479 K. After that, this rate suddenly slowed down and it was only 3.5 mg/min at 503 K. Following that, a small increase to 3.9 mg/min and then a serious decrease took place as the temperature increases. At low temperatures, lignite behaved as not as reactive as hazelnut shell. Although some devolatilization was observed at low temperatures, the rates were not higher than 1.5 mg/min. The Rmax value was determined as 1.75 mg/min at 580 K. At higher temperatures, both samples showed the same behavior that relatively lower rates of mass losses continued until the final temperature of the experiment (Fig. 3). Lignin, which is a more stable compound compared to holocellulose in biomass, gave rise to these mass losses at high temperatures. Conversely, surface oxidation of fixed carbon in lignite, and transformations of some ash forming mineral species can be attributed to the mass losses in Elbistan lignite at high temperatures. The rates of mass losses from lignite at the high temperature region were generally higher than those from biomass. This can be explained by the fact that lignite has a greater fixed carbon content than biomass. On the other hand, final burnouts were determined as 82.1% and 97.5% for lignite and biomass on the raw mass of sample, respectively. Volatile matter/fixed carbon ratio is generally higher than 4.0 for biomass, while it is considerably lower for coal (Tillman, 2000). This ratio in this study was found as 10.4 and 2.9 for hazelnut shell and lignite samples, respectively. This predicts the tendency of the samples toward devolatilization and gaseous phase combustion behavior. In this context, high mass losses from the biomass at low temperatures are convenient with the volatile matter/fixed carbon ratio in this sample. On the other hand, biomass comprises mainly cellulosics, hemicellulosics, lignin, and extractives. AOH groups are the principal functional groups in most biomass species (Tillman, 2000). These groups have an important role on the high thermal reactivity of biomass. However, oxygen in the functional groups of low rank coals is mainly in the functional groups of ACOOH, AOCH3, and AOH (Yaman and Kucukbayrak, 2000; Yaman et al., 2000). This may be another reason for the differences between the thermal reactivity of the samples. Besides, concentrations of some elements, especially sodium and potassium that lead to high reactivity during combustion, are usually higher in biomass.
2081
3.3. Combustion properties of the blends Table 2 represents the comparison of final burnout, Rmax, and TR-max values. Mass losses during the moisture removal stage were not taken into account. Since biomass is the minor constituent in the blends, the thermal properties of the lignite dominated the characteristics of the blends, especially in the biomass-lean blends. In the cases of the 10 wt% and 20 wt% blends, the effects of biomass addition could be seen very well. The highest burnout among the blend samples did not exceed the value measured for biomass itself. However, the lowest burnout obtained for the 10 wt% blend was even lower than the value measured for lignite. Table 2 reveals that the addition of biomass into lignite increased the extents of Rmax and decreased the values of TR-max depending on the amount of biomass in the blend. The ‘‘easy-to-ignite’’ property of biomass is responsible for these trends. In fact, biomass decreases the ignition point of the blend and its contribution to the volatile formation enhances the burning rate. On the other hand, the influence of biomass on the burnout yields is seen to be more complicated. In case of small additions of biomass, burnout yields become higher. But, addition of 10 wt% or 20 wt% of biomass did not cause the expected increasing effect on burnouts. Besides, it was seen that a dramatic reduction to 76.5% in burnout occurred when biomass is blended with lignite in the ratio of 10 wt%. In order to investigate this complexity of biomass addition on burnouts, the effect of temperature was evaluated. For this purpose, the temperature was varied from 423 K to 1173 K by increments of 50 K. The burnouts at specific temperatures are given in Table 3. From the data in Table 3, it can be said that the experimental burnouts for all of the blends except for 10 wt% lay between the burnout values of the parent values at the investigated temperatures. Burnout curves of the blends drawn versus temperature would be between the curves of the parent samples, and they would be mostly overlapped. Hence, in order to highlight the behavior of the blend of 10 wt%, only this blend was incorporated to Fig. 4. Theoretical values of burnouts for the blends were calculated from sum of the fractional contributions of the experimental values of biomass and lignite in the blends. Table 2 Comparison of some thermal data
Lignite Biomass 2% 4% 6% 8% 10% 20%
Final burnout (%)
Rmax (mg/min)
TR-max (K)
82.1 97.5 89.3 88.6 88.3 89.3 76.5 86.8
1.75 5.50 2.30 2.40 2.50 3.10 3.20 3.50
580 479 493 483 480 473 470 463
2082
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Table 3 Experimental burnout values versus temperature Temperature (K)
Lignite
Biomass
2%
4%
6%
8%
10%
20%
423 473 523 573 623 673 723 773 823 873 923 973 1023 1073 1123 1173
17.00 25.70 34.60 40.78 44.39 48.01 50.76 52.98 55.21 58.42 61.22 65.93 70.66 74.87 78.57 82.14
15.74 22.59 55.33 64.12 68.02 71.32 73.55 74.96 77.16 79.40 81.85 84.52 87.13 90.61 94.29 97.46
19.14 29.16 39.10 44.66 50.25 54.52 58.01 61.07 64.38 67.05 70.74 75.40 79.96 84.05 87.83 89.31
18.52 27.99 37.83 45.04 49.87 53.44 57.38 60.17 62.98 65.65 69.08 73.28 77.86 81.51 85.50 88.55
17.6 26.09 37.78 45.92 50.65 54.64 58.93 61.14 63.52 66.84 69.74 73.72 77.68 81.38 85.20 88.26
18.34 29.95 40.10 45.68 50.15 54.16 57.61 60.91 63.17 66.65 70.81 75.00 77.41 82.61 86.55 89.34
13.67 27.96 33.54 38.48 42.22 45.57 47.72 49.37 51.77 54.43 57.73 61.27 64.98 68.61 72.91 76.46
18.32 35.11 40.84 45.36 49.11 52.16 54.20 56.40 59.29 62.09 65.48 69.26 74.43 78.83 83.54 86.77
100
Burn-out (%)
80
60
40
10% Lignite Biomass
20
0 200
400
600
800
1000
1200
Temperature (K) Fig. 4. Burnout yields of the parent samples and 10% blend (on raw mass of fuel).
The following equation was used to calculate the deviations of experimental burnouts from the theoretical ones, which are on the basis of raw mass of fuel: D¼
ðBexp Btheo Þ 100 Btheo
ð2Þ
where D is the Deviation (%), Bexp is the experimental burnout (%) and Btheo is the theoretical burnout (%). For the blends of 2–8 wt%, deviations were determined as positive values, indicating higher burnout values than expected. This shows that the addition of biomass into lignite in the ratios up to 8 wt% caused an increase in the burnout values detected from the co-combustion experiments at every temperature. Compared to Elbistan lignite, hazelnut shell has higher volatile matter, and lower fixed carbon and ash contents. Therefore, it is possible that the addition of hazelnut shell into Elbistan lignite affects the combustion system in such a way that volatilization and gaseous phase combustion become more important compared to the combustion of coal alone. Hence, higher burnout values were
determined as a result of the changing characteristics of the fuel. Similarly, Xiang-guo et al. (2006) determined that the co-combustion of coal and waste tires can improve the burnout efficiency of high ash coal to a certain degree. On the other hand, almost all of the deviation values in Table 4 for the 10 wt% blend indicate an opposite trend, so that negative deviation values were detected for this blend sample. Burnout values which are lower than the theoretical ones were found, and this indicates the different characteristics of this blending ratio. This may be because of some interactions which took place between the constituents of both fuels, and consequently much more mass remained without altering to gaseous form. The fact that the burnout values were different increasingly or decreasingly from the theoretically expected values for the blend samples is a result of the synergistic interaction between the two different types of fuels. Regardless of the real reason, it can be said that some synergistic effects are in effect during the co-combustion of Elbistan lignite and hazelnut shell in terms of the burnout yields. Table 4 Deviations of the experimental burnouts from the theoretical values Temperature (K)
2%
4%
6%
8%
10%
20%
423 473 523 573 623 673 723 773 823 873 923 973 1023 1073 1123 1173
12.76 13.74 11.67 8.28 12.01 12.47 13.27 14.32 15.69 13.95 14.78 13.72 12.64 11.79 11.34 8.32
9.27 9.44 6.78 7.97 10.00 9.19 11.05 11.72 12.29 10.78 11.34 9.91 9.17 7.96 7.96 7.01
3.99 2.26 5.40 8.87 10.57 10.59 13.05 12.60 12.37 12.00 11.66 9.96 8.42 7.34 7.15 6.26
8.53 17.68 10.60 7.11 8.36 8.59 9.56 11.27 10.89 10.90 12.63 11.25 7.55 8.51 8.42 7.17
18.99 10.13 8.54 10.75 9.70 9.48 10.03 10.53 9.82 10.06 8.77 9.62 10.13 10.25 9.02 8.62
9.39 40.00 5.40 0.19 0.01 0.97 2.02 1.70 0.52 0.84 0.21 0.56 0.64 1.04 2.23 1.84
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084 Table 5 Apparent activation energies Sample
Apparent Ea (kJ/mol)
Lignite Biomass 2% 4% 6% 8% 10% 20%
84.3 60.4 68.9 68.5 67.6 67.1 66.8 63.4
In case of the 20 wt% blend, although signs of the deviations may change, their values vary in a narrow range in general. In other words, the burnout values which are near to the expected values were found only for this blend. From this point of the view, additive behavior is more favorable for this blending ratio. Similar behavior was also observed by Folgueras et al. (2003) for the co-combustion of coal and sewage sludge. The above mentioned interactions in the burnouts indicate that if biomass is added to lignite in small quantities, it facilitates the combustion of coal and so higher burnouts can be obtained. Similarly, Laursen and Grace (2002) reported that satisfactory burnout levels can be achieved when hog fuel and coal were burned together. However, higher ratios of biomass in the blend did not provide the same behavior. Addition of biomass into lignite in the ratio of 10 wt% apparently led to decrease in the burnouts of the blend at all of the temperatures except for 473 K. This may be from the physical and chemical restrictions during the process. Table 5 shows the apparent activation energy levels for the samples investigated. The values in Table 5 indicate that the addition of hazelnut shell to Elbistan lignite facilitates the combustion process in general, and consequently lower apparent activation energies are sufficient to conduct this process. The variation range of the apparent activation energies found in this study agreed well with the literature (Folgueras et al., 2003). 4. Conclusions Hazelnut shell and Elbistan lignite, which are important energy sources in Turkey, have important differences in their physical, chemical, and thermal properties. Blending these two fuel samples may have the potential for compensating the negative aspects of each other. However, the thermal reactivity of each sample during combustion is highly distinctive. The addition of hazelnut shell into Elbistan lignite affects the combustion system in such a way that volatilization and gaseous phase combustion become more important compared to the combustion of coal alone. Burnouts obtained for the blend samples were not always at the values which are calculated from the fractions of the experimental values of the parent samples. The addition of
2083
hazelnut shell up to 8 wt% into Elbistan lignite provided higher burnouts than the expected theoretical ones. However, the addition of 10 wt% biomass played a reducing role on the burnouts. Burnout characteristics of the blends up to 10 wt% during co-combustion can be concluded to show synergistic effect, whereas additive effects are more favorable in the case of the blend having a biomass content of 20 wt%. On the other hand, apparent activation energy, maximum burning rate, and temperature of the maximum burning rate can be concluded to follow the additive behavior for the blend samples. References Amand, L.E., Leckner, B., Eskilsson, D., Tullin, C., 2006. Deposits on heat transfer tubes during co-combustion of biofuels and sewage sludge. Fuel 85, 1313–1322. Coats, A.W., Redfern, J.P., 1964. Kinetic parameters from thermogravimetric data. Nature 201, 68–69. Demirbas, A., 2002. Fuel characteristics of olive husk and walnut, hazelnut, sunflower, and almond shells. Energy Sources 24, 215– 221. Diaz-Somoano, M., Unterberger, S., Hein, K.R.G., 2006. Prediction of trace element volatility during co-combustion processes. Fuel 85, 1087–1093. Elled, A.L., Amand, L.E., Leckner, B., Anderson, B.A., 2006. Influence of phosphorus on sulphur capture during co-firing of sewage sludge with wood or bark in a fluidized bed. Fuel 85, 1671–1678. Ferrer, E., Aho, M., Silvennoinen, J., Nurminen, R.V., 2005. Fluidized bed combustion of refuse-derived fuel in presence of protective coal ash. Fuel Processing Technology 87, 33–44. Folgueras, M.B., Diaz, R.M., Xiberta, J., Prieto, I., 2003. Thermogravimetric analysis of the co-combustion of coal and sewage sludge. Fuel 82, 2051–2055. Gulyurtlu, I., Boavida, D., Abelha, P., Lopes, M.H., Cabrita, I., 2005. Cocombustion of coal and meat and bone meal. Fuel 84, 2137–2148. Hartmann, D., Kaltschmitt, M., 1999. Electricity generation from solid biomass via co-combustion with coal-energy and emission balances from a German case study. Biomass and Bioenergy 16, 397–406. Haykiri-Acma, H., Yaman, S., Kucukbayrak, S., 2006. Effect of heating rate on the pyrolysis yields of rapeseed. Renewable Energy 31, 803– 810. Karayigit, A.I., Gayer, R.A., Querol, X., Onacak, T., 2000. Contents of major and trace elements in feed coals from Turkish coal-fired power plants. International Journal of Coal Geology 44, 169–184. Klass, D.L., 1998. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego. Laursen, K., Grace, J.R., 2002. Some implications of co-combustion of biomass and coal in a fluidized bed boiler. Fuel Processing Technology 76, 77–89. Martin, C., Villamanan, M.A., Chamorro, C.R., Otero, J., Cabanillas, A., Segovia, J.J., 2006. Low-grade coal and biomass co-combustion on fluidized bed: exergy analysis. Energy 31, 330–344. Pronobis, M., 2006. The influence of biomass co-combustion on boiler fouling and efficiency. Fuel 85, 474–480. Raghunathan, K., Gullett, B.K., 1996. Role of sulfur in reducing PCDD and PCDF formation. Environmental Science and Technology 30, 1827–1834. Sanchez-Hervas, J.M., Armesto, L., Ruiz-Martinez, E., Otero-Ruiz, J., Pandelova, M., Schramm, K.W., 2005. PCDD/PCDF emissions from co-combustion of coal and PVC in a bubbling fluidized bed boiler. Fuel 84, 2149–2157. Steenari, B.M., Lindqvist, O., 1999. Fly ash characteristics in cocombustion of wood with coal, oil or peat. Fuel 78, 479–488. TAPPI T203 om-88, 1988. Alpha-, beta- and gamma-cellulose in pulp.
2084
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Theis, M., Skrifvars, B.J., Hupa, M., Tran, H., 2006. Fouling tendency of ash resulting from burning mixtures of biofuels. Part 1: Deposition rates. Fuel 85, 1125–1130. Tillman, D.A., 2000. Biomass cofiring: the technology, the experience, the combustion consequences. Biomass and Bioenergy 19, 365–384. Topuz, A., Topakci, M., Canakci, M., Akinci, I., Ozdemir, F., 2005. Physical and nutritional properties of four orange varieties. Journal of Food Engineering 66, 519–523. Ural, S., 2005. Comparison of fly ash properties from Afsin-Elbistan Coal Basin Turkey. Journal of Hazardous Materials 119, 85–92. van Krevelen, D.W., 1993. Coal: Typology–Physics–Chemistry-Constitution. Elsevier Science Ltd., Amsterdam.
van Soest, P.J., 1963. Use of detergents in the analysis of fibrous feeds. II. A rapid method for the determination of fiber and lignin. Journal of Association of Official Analytical Chemists 46, 830. Xiang-guo, L., Bao-guo, M., Li, X., Zhen-wu, H., Xin-gang, W., 2006. Thermogravimetric analysis of the co-combustion of the blends with high ash and waste tyres. Thermochimica Acta 441, 79–83. Yaman, S., Kucukbayrak, S., 2000. Influence of oxidation in aqueous medium on the structure and properties of lignites. Fuel 79, 777–783. Yaman, S., Karatepe, N., Kucukbayrak, S., 2000. Influence of wet oxidation on the surface area and the porosity of some lignites. Fuel 79, 1017–1022.
Waste Management 28 (2008) 2077–2084 www.elsevier.com/locate/wasman
Effect of co-combustion on the burnout of lignite/biomass blends: A Turkish case study H. Haykiri-Acma, S. Yaman
*
Istanbul Technical University, Chemical and Metallurgical Engineering Faculty, Chemical Engineering Department, 34469 Maslak, Istanbul, Turkey Accepted 22 August 2007 Available online 26 October 2007
Abstract Co-combustion of Turkish Elbistan lignite and woody shells of hazelnut was performed in a TGA up to 1173 K with a heating rate of 20 K/min. SEM images of each fuel revealed the differences in their physical appearances. Hazelnut shell was blended with lignite in the range of 2–20 wt% to observe the co-combustion properties. Maximum burning rates (Rmax), temperatures of the maximum burning rates (TR-max), and the final burnout values of the parent samples and the blends were compared. The results were interpreted considering lignite properties and the major biomass ingredients such as cellulosics, hemicellulosics, and lignin. Deviations between the theoretical and experimental burnout values were evaluated at various temperatures. Burnout characteristics of the blends up to 10 wt% were concluded to have a synergistic effect so the addition of hazelnut shell up to 8 wt% provided higher burnouts than the expected theoretical ones, whereas addition of as much as 10 wt% led to a decrease in the burnout. However, the additive effects were more favorable for the blend having a biomass content of 20 wt%. Apparent activation energy, Rmax, and TR-max, were found to follow the additive behavior for the blend samples. Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction Co-combustion is a common practice and an effective method for waste disposal, not only can significant energy recovery be obtained from waste materials but also a marked reduction in the volume of the wastes is realized. There are several reasons to blend waste materials with coal or with other types of fuels prior to burning. In particular, the ash deposition and fouling problems on hot surfaces, which are commonly encountered in combustion of biomass, can be diminished or eliminated by means of burning of biomass/coal blends. On the other hand, the ash fraction of various biomass species is usually rich in alkaline and alkaline earth elements, providing an important potential to capture the SO2 being released from the combustion of high sulfur coals (Diaz-Somoano et al., 2006; Martin et al., 2006).
*
Corresponding author. Tel.: +90 212 2856873; fax: +90 212 2852925. E-mail address: [email protected] (S. Yaman).
0956-053X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2007.08.028
Since fresh biomass fixes solar radiation and CO2 in the form of lignocellulosics during photosynthesis, the emitted CO2 from combustion of these lignocellulosic materials is regarded not to have any net contribution to the CO2 accumulation in atmosphere. Hence, compared to combustion of fossil fuels alone, usage of biomass-blended fuels is less harmful in terms of greenhouse effect (Klass, 1998). Besides, a fuel is accepted as non-corrosive if its S/Cl molar ratio is greater than 4, whereas fuels with S/Cl molar ratios that are lower than 2 lead to serious corrosion in materials (Ferrer et al., 2005). This ratio can be adjusted by coexisting biomass and coal in the combustion medium. Co-combustion of municipal solid wastes (MSW) and agricultural residues offers the possibility of reducing the need for landfill sites owing to the significant reductions in the amounts of the waste materials, and improvements in the ash properties of the blended fuels compared to that of the parent waste materials (Steenari and Lindqvist, 1999). In the case of refuse derived fuel (RDF) combustion, polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo-p-furans (PCDFs) emissions are formed,
2078
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
causing serious environmental and health concerns. However, co-firing of RDF with coal is reported to inhibit the formation of these dangerous emissions considerably (Ferrer et al., 2005). Also, Raghunathan and Gullett (1996) showed that SO2 injection was evidently capable of preventing the formation of PCDD/F. Likewise, the presence of coal during co-firing also provides some reduction in the emissions of PCDD/Fs resulting from the combustion of PVC (Sanchez-Hervas et al., 2005). Combustion of blended straw or waste wood with coal was reported to have reasonably low environmental effects (Hartmann and Kaltschmitt, 1999). On the other hand, characteristics of ashes released from the co-combustion of meat and bone meal (MBM) and coal was seen as appropriate to landfill without environmental risk (Gulyurtlu et al., 2005). The composition of ash from biomass materials is generally different from ash from coal (van Krevelen, 1993). Alkalinity is usually higher in biomass ashes. The typical ratio of base/acid is calculated as follows: Base Fe2 O3 þ CaO þ MgO þ Na2 O þ K2 O þ P2 O5 ¼ Acid SiO2 þ Al2 O3 þ TiO2
ð1Þ
This ratio for biomass often exceeds 1.0 and may be beyond 2.0, whereas it is usually lower than 1.0 in for coal. The base/acid ratio is an indication of fuel performance in association with problematic deposit formations (Tillman, 2000; Pronobis, 2006). Potassium is the key element in most biomass species, leading to the formation of the undesired depositions and fouling on the heat exchanger surfaces. However, the sulfur content of the fuel that biomass is blended with may react with potassium to form less harmful products. Besides, potassium in a gaseous form may react with alumina-silicates coming from the coal in the blend (Amand et al., 2006). Combustion of low rank coals with biomass in power stations is reported to reduce the emissions of greenhouse gases, NOx, and SOx (Martin et al., 2006). It is a common practice to fire various organic wastes with coal in order to reduce the volume of dangerous waste, heat recovery, and removal of toxic organics (Elled et al., 2006). In various blends with straw, peat behaved as a cleaning agent with respect to deposition rate during co-firing. This was attributed to the alumina-silicate rich character of peat ash (Theis et al., 2006). Hazelnut production is of great importance in the Turkish agricultural sector so that Turkey generates about 80% of the world’s hazelnuts (Topuz et al., 2005). Consequently, several 100,000 tonnes of hazelnut wastes are renewably formed every year. The high calorific value of hazelnut shells makes it a potential energy source. This fuel is generally used in domestic heating or in small-scale combustion systems near to the cultivation areas. Ash from hazelnut shell contains a K2O content of 30.4 wt% (db), which is almost 50% of the sum of the basic components, and this leads to a base/acid ratio of 2.03 which is somewhat high
compared to coal (Demirbas, 2002). Such a high concentration of K2O is the main drawback encountered during the combustion of this potential fuel in commercial combustors. On the other hand, Elbistan lignite is the most important national primary energy resource in Turkey with a reserve of 3.4 billion metric tonnes. However, it has quite a low calorific value, and high sulfur and high ash contents. Therefore, effective measures must be taken in order to reduce the environmental risks resulting from the combustion of Elbistan lignite. Nevertheless, the K2O content of Elbistan lignite is very low, and it is less than 1 wt% even in its fly ash (Ural, 2005). Calcite is the key mineral species in Elbistan lignite; 27.4 wt% of dry coal is calcite. Other important minerals are quartz (4.8%), illite (2.6%), pyrite (1.6%), kaolinite (0.5%), and aragonite (0.5%). Calcium consists of 11.1 wt% of dry Elbistan lignite (Karayigit et al., 2000). Phosphor is regarded to inhibit the retention of sulphur by calcium compounds (Elled et al., 2006). In this context, since the phosphor content of hazelnut shell is not as high as in RDF or MBM, this effect is not expected to occur during co-combustion of hazelnut shell and high sulphur coals. For the outlined reasons, Elbistan lignite and hazelnut shell seem to have important potential to compensate the disadvantages coming from each other. Biomass may contribute to the calorific value of the low rank-low quality coal, and lignite may behave as a cleaning agent in terms of fouling or depositions on convective heating surfaces resulting from biomass. However, apparent differences in the chemical and physical properties of both fuels may cause some inconveniences in the combustion systems due to the different combustion characteristics of each fuel. Furthermore, burnout characteristics of their blends must be taken into account as another challenge. Hence, the aim of this paper is to compare the combustion properties of each fuel under the same slow combustion conditions, and to investigate the co-combustion characteristics and the burnouts of their blends. 2. Experimental In the experiments, woody shells of Turkish hazelnuts were used as the biomass material. Elbistan lignite was selected as the coal sample. Both samples were milled and sieved to a particle size of <0.250 mm. The proximate analysis and the gross calorific value measurements of the samples were conducted according to ASTM standards. Ultimate analyses of the samples were carried out using an EuroEA3000 model elemental analyzer. Major organic components in hazelnut shell were determined by analytical manner. For this purpose, extractives were determined applying the benzene–ethyl alcohol extraction procedure according to ASTM D1105. The bulk remaining after this procedure was treated with the mixtures
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
of NaClO2 and acetic acid to isolate the holocellulose (cellulosics + hemicellulosics) content. The lignin content was determined by the van Soest method (1963); a-cellulose content was determined according to TAPPI T203 standard (1988). Scanning electron microscopy (SEM) tests were performed in a Joelä Model JSM-T330 operated at 25 kV and linked with an energy dispersive (EDS) attachment to study the physical appearances of the fuel samples. Lignite and hazelnut shell were mixed to obtain various blends having hazelnut shell in the ratios of 2, 4, 6, 8, 10, and 20 wt%. Slow combustion of both parent fuels and their blends were performed in a Shimadzu TG 41 Thermogravimetric Analyzer (TGA) with a cylindrical alumina crucible. In these experiments, 40 mg of sample was heated from ambient to 1173 K by a constant heating rate of 20 K/min under dry air flow of 40 mL/min. Then, the sample was held at 1173 K until no further mass loss was detected. Derivative thermogravimetric analysis (DTG) profiles, showing the burning rates of the samples, were derived from the TGA data; and the Coats–Redfern method (1964) was applied to calculate the apparent activation energy values during the process. All experiments were repeated, and the mean values were used provided that the deviations were within 5%. 3. Results and discussion 3.1. Differences between the properties of the samples Table 1 presents the results of the proximate and ultimate analyses, calorific values of both samples, and major biomass components of hazelnut shell. Elbistan lignite is a low grade coal. Low fixed carbon and high ash contents are its typical characteristics. Besides, a sulfur content of as much as 3.6 wt% (daf) is seriously high,
2079
and effective measures must be taken in order to get rid of the problems resulting from SO2 emissions. On the other hand, its calorific value is considerably low. Having high sulfur, high ash, low fixed carbon, and low calorific value makes Elbistan lignite extremely problematic. Unlike Elbistan lignite, hazelnut shell has a relatively higher calorific value, lower ash content, and negligible sulfur that could not be detected by the analyzer used. More than half of the organic part of hazelnut shell is comprised from lignin. This is consistent with the woody structure of this sample; a-cellulose is the richest constituent in the holocellulose. However, the oxygen content of hazelnut was higher than that of Elbistan lignite. This indicates that this biomass material would have higher thermal reactivity than lignite. SEM images of the samples having particle sizes of <0.250 mm are illustrated in Fig. 1. According to these images, it can be said that there are fundamental differences in the physical appearances of both samples. These differences can be summarized as follows: Lignite is very rich in small particles in such a way that some particles with the sizes of as low as several microns can be easily noticed. This implies the fragile structure and the tendency toward dust formation of Elbistan lignite during milling. In fact, this is an important problem in the mining of this lignite. Whereas, small particles are relatively less in the case of hazelnut shell, owing to its fibrous structure. Lignite particles are seen to have spherical shape in general, and the borders of the particles are relatively regular. Although, hazelnut particles have a distinctly amorphous shape, they may be assumed to be rather cylindrical. In addition to the shape, the surface areas of the different size particles show fundamental differences for each sample. Some bright parts are available on the surfaces of the lignite particles, which may be attributed to the reflection of the mineral phases found in the lignite.
Table 1 Analyses results of hazelnut shell and Elbistan lignite Analysis
Hazelnut shell
Lignite
Moisture (%)
11.8
45.0
Volatile matter (%, db) Fixed carbon (%, db) Ash (%, db)
87.3 8.4 4.3
50.5 17.2 32.3
C (%, daf) H (%, daf) N (%, daf) S (%, daf) Oa (%, daf)
46.3 5.8 0.4 <0.05 47.5
53.2 5.5 1.7 3.6 36.0
Gross calorific value (kcal/kg, db) Extractives (%, db) Holocellulose (%, db) Lignin (%, db) a-Cellulose (%, db) a
Calculated by difference.
4358 6.2 38.6 51.5 22.9
2400 n/a n/a n/a n/a
3.2. Combustion properties of the parent samples TGA and DTG burning profiles of the samples are given in Figs. 2 and 3, respectively. Drying, thermal decomposition, volatile release, combustion of volatiles, char gasification, and char oxidation stages caused mass losses from the solid matrices of the samples. Before the initiation of the burning, some increase in mass could be observed due to the chemisorption of oxygen. Folgueras et al. (2003) also reported a similar increase in the mass of the sample during such an experiment. Maximum burning rates (Rmax), temperatures of the maximum burning rates (TR-max), and final burnout values were obtained from these profiles, and considered to compare the thermal behaviors of the samples under investigated combustion conditions.
2080
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Fig. 1. SEM images of the Elbistan lignite and hazelnut shell samples.
dm/dt (mg/min)
m (mg)
30
20
10
0 200
6
Lignite Biomass
40
400
600
800
1000
Temperature (K) Fig. 2. TGA curves of the samples.
1200
Lignite Biomass
4
2
0 200
400
600
800
1000
Temperature (K) Fig. 3. DTG profiles of the samples.
1200
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Fig. 2 shows that mass losses occurred from the samples upon heating. In spite of the relatively low temperatures, important structural decompositions occurred, especially in the hazelnut shell just after moisture release, and consequently marked devolatilization yields were detected. Holocellulosics in the hazelnut shell is responsible for these mass losses due to their high reactivity and weak structures (Haykiri-Acma et al., 2006). After that, cellulosics contributed to the formation of volatiles. Ignition of combustible volatiles augmented the mass losses from the sample. Thus, the rate of mass loss on Fig. 3 reached 5.50 mg/min at 479 K. After that, this rate suddenly slowed down and it was only 3.5 mg/min at 503 K. Following that, a small increase to 3.9 mg/min and then a serious decrease took place as the temperature increases. At low temperatures, lignite behaved as not as reactive as hazelnut shell. Although some devolatilization was observed at low temperatures, the rates were not higher than 1.5 mg/min. The Rmax value was determined as 1.75 mg/min at 580 K. At higher temperatures, both samples showed the same behavior that relatively lower rates of mass losses continued until the final temperature of the experiment (Fig. 3). Lignin, which is a more stable compound compared to holocellulose in biomass, gave rise to these mass losses at high temperatures. Conversely, surface oxidation of fixed carbon in lignite, and transformations of some ash forming mineral species can be attributed to the mass losses in Elbistan lignite at high temperatures. The rates of mass losses from lignite at the high temperature region were generally higher than those from biomass. This can be explained by the fact that lignite has a greater fixed carbon content than biomass. On the other hand, final burnouts were determined as 82.1% and 97.5% for lignite and biomass on the raw mass of sample, respectively. Volatile matter/fixed carbon ratio is generally higher than 4.0 for biomass, while it is considerably lower for coal (Tillman, 2000). This ratio in this study was found as 10.4 and 2.9 for hazelnut shell and lignite samples, respectively. This predicts the tendency of the samples toward devolatilization and gaseous phase combustion behavior. In this context, high mass losses from the biomass at low temperatures are convenient with the volatile matter/fixed carbon ratio in this sample. On the other hand, biomass comprises mainly cellulosics, hemicellulosics, lignin, and extractives. AOH groups are the principal functional groups in most biomass species (Tillman, 2000). These groups have an important role on the high thermal reactivity of biomass. However, oxygen in the functional groups of low rank coals is mainly in the functional groups of ACOOH, AOCH3, and AOH (Yaman and Kucukbayrak, 2000; Yaman et al., 2000). This may be another reason for the differences between the thermal reactivity of the samples. Besides, concentrations of some elements, especially sodium and potassium that lead to high reactivity during combustion, are usually higher in biomass.
2081
3.3. Combustion properties of the blends Table 2 represents the comparison of final burnout, Rmax, and TR-max values. Mass losses during the moisture removal stage were not taken into account. Since biomass is the minor constituent in the blends, the thermal properties of the lignite dominated the characteristics of the blends, especially in the biomass-lean blends. In the cases of the 10 wt% and 20 wt% blends, the effects of biomass addition could be seen very well. The highest burnout among the blend samples did not exceed the value measured for biomass itself. However, the lowest burnout obtained for the 10 wt% blend was even lower than the value measured for lignite. Table 2 reveals that the addition of biomass into lignite increased the extents of Rmax and decreased the values of TR-max depending on the amount of biomass in the blend. The ‘‘easy-to-ignite’’ property of biomass is responsible for these trends. In fact, biomass decreases the ignition point of the blend and its contribution to the volatile formation enhances the burning rate. On the other hand, the influence of biomass on the burnout yields is seen to be more complicated. In case of small additions of biomass, burnout yields become higher. But, addition of 10 wt% or 20 wt% of biomass did not cause the expected increasing effect on burnouts. Besides, it was seen that a dramatic reduction to 76.5% in burnout occurred when biomass is blended with lignite in the ratio of 10 wt%. In order to investigate this complexity of biomass addition on burnouts, the effect of temperature was evaluated. For this purpose, the temperature was varied from 423 K to 1173 K by increments of 50 K. The burnouts at specific temperatures are given in Table 3. From the data in Table 3, it can be said that the experimental burnouts for all of the blends except for 10 wt% lay between the burnout values of the parent values at the investigated temperatures. Burnout curves of the blends drawn versus temperature would be between the curves of the parent samples, and they would be mostly overlapped. Hence, in order to highlight the behavior of the blend of 10 wt%, only this blend was incorporated to Fig. 4. Theoretical values of burnouts for the blends were calculated from sum of the fractional contributions of the experimental values of biomass and lignite in the blends. Table 2 Comparison of some thermal data
Lignite Biomass 2% 4% 6% 8% 10% 20%
Final burnout (%)
Rmax (mg/min)
TR-max (K)
82.1 97.5 89.3 88.6 88.3 89.3 76.5 86.8
1.75 5.50 2.30 2.40 2.50 3.10 3.20 3.50
580 479 493 483 480 473 470 463
2082
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Table 3 Experimental burnout values versus temperature Temperature (K)
Lignite
Biomass
2%
4%
6%
8%
10%
20%
423 473 523 573 623 673 723 773 823 873 923 973 1023 1073 1123 1173
17.00 25.70 34.60 40.78 44.39 48.01 50.76 52.98 55.21 58.42 61.22 65.93 70.66 74.87 78.57 82.14
15.74 22.59 55.33 64.12 68.02 71.32 73.55 74.96 77.16 79.40 81.85 84.52 87.13 90.61 94.29 97.46
19.14 29.16 39.10 44.66 50.25 54.52 58.01 61.07 64.38 67.05 70.74 75.40 79.96 84.05 87.83 89.31
18.52 27.99 37.83 45.04 49.87 53.44 57.38 60.17 62.98 65.65 69.08 73.28 77.86 81.51 85.50 88.55
17.6 26.09 37.78 45.92 50.65 54.64 58.93 61.14 63.52 66.84 69.74 73.72 77.68 81.38 85.20 88.26
18.34 29.95 40.10 45.68 50.15 54.16 57.61 60.91 63.17 66.65 70.81 75.00 77.41 82.61 86.55 89.34
13.67 27.96 33.54 38.48 42.22 45.57 47.72 49.37 51.77 54.43 57.73 61.27 64.98 68.61 72.91 76.46
18.32 35.11 40.84 45.36 49.11 52.16 54.20 56.40 59.29 62.09 65.48 69.26 74.43 78.83 83.54 86.77
100
Burn-out (%)
80
60
40
10% Lignite Biomass
20
0 200
400
600
800
1000
1200
Temperature (K) Fig. 4. Burnout yields of the parent samples and 10% blend (on raw mass of fuel).
The following equation was used to calculate the deviations of experimental burnouts from the theoretical ones, which are on the basis of raw mass of fuel: D¼
ðBexp Btheo Þ 100 Btheo
ð2Þ
where D is the Deviation (%), Bexp is the experimental burnout (%) and Btheo is the theoretical burnout (%). For the blends of 2–8 wt%, deviations were determined as positive values, indicating higher burnout values than expected. This shows that the addition of biomass into lignite in the ratios up to 8 wt% caused an increase in the burnout values detected from the co-combustion experiments at every temperature. Compared to Elbistan lignite, hazelnut shell has higher volatile matter, and lower fixed carbon and ash contents. Therefore, it is possible that the addition of hazelnut shell into Elbistan lignite affects the combustion system in such a way that volatilization and gaseous phase combustion become more important compared to the combustion of coal alone. Hence, higher burnout values were
determined as a result of the changing characteristics of the fuel. Similarly, Xiang-guo et al. (2006) determined that the co-combustion of coal and waste tires can improve the burnout efficiency of high ash coal to a certain degree. On the other hand, almost all of the deviation values in Table 4 for the 10 wt% blend indicate an opposite trend, so that negative deviation values were detected for this blend sample. Burnout values which are lower than the theoretical ones were found, and this indicates the different characteristics of this blending ratio. This may be because of some interactions which took place between the constituents of both fuels, and consequently much more mass remained without altering to gaseous form. The fact that the burnout values were different increasingly or decreasingly from the theoretically expected values for the blend samples is a result of the synergistic interaction between the two different types of fuels. Regardless of the real reason, it can be said that some synergistic effects are in effect during the co-combustion of Elbistan lignite and hazelnut shell in terms of the burnout yields. Table 4 Deviations of the experimental burnouts from the theoretical values Temperature (K)
2%
4%
6%
8%
10%
20%
423 473 523 573 623 673 723 773 823 873 923 973 1023 1073 1123 1173
12.76 13.74 11.67 8.28 12.01 12.47 13.27 14.32 15.69 13.95 14.78 13.72 12.64 11.79 11.34 8.32
9.27 9.44 6.78 7.97 10.00 9.19 11.05 11.72 12.29 10.78 11.34 9.91 9.17 7.96 7.96 7.01
3.99 2.26 5.40 8.87 10.57 10.59 13.05 12.60 12.37 12.00 11.66 9.96 8.42 7.34 7.15 6.26
8.53 17.68 10.60 7.11 8.36 8.59 9.56 11.27 10.89 10.90 12.63 11.25 7.55 8.51 8.42 7.17
18.99 10.13 8.54 10.75 9.70 9.48 10.03 10.53 9.82 10.06 8.77 9.62 10.13 10.25 9.02 8.62
9.39 40.00 5.40 0.19 0.01 0.97 2.02 1.70 0.52 0.84 0.21 0.56 0.64 1.04 2.23 1.84
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084 Table 5 Apparent activation energies Sample
Apparent Ea (kJ/mol)
Lignite Biomass 2% 4% 6% 8% 10% 20%
84.3 60.4 68.9 68.5 67.6 67.1 66.8 63.4
In case of the 20 wt% blend, although signs of the deviations may change, their values vary in a narrow range in general. In other words, the burnout values which are near to the expected values were found only for this blend. From this point of the view, additive behavior is more favorable for this blending ratio. Similar behavior was also observed by Folgueras et al. (2003) for the co-combustion of coal and sewage sludge. The above mentioned interactions in the burnouts indicate that if biomass is added to lignite in small quantities, it facilitates the combustion of coal and so higher burnouts can be obtained. Similarly, Laursen and Grace (2002) reported that satisfactory burnout levels can be achieved when hog fuel and coal were burned together. However, higher ratios of biomass in the blend did not provide the same behavior. Addition of biomass into lignite in the ratio of 10 wt% apparently led to decrease in the burnouts of the blend at all of the temperatures except for 473 K. This may be from the physical and chemical restrictions during the process. Table 5 shows the apparent activation energy levels for the samples investigated. The values in Table 5 indicate that the addition of hazelnut shell to Elbistan lignite facilitates the combustion process in general, and consequently lower apparent activation energies are sufficient to conduct this process. The variation range of the apparent activation energies found in this study agreed well with the literature (Folgueras et al., 2003). 4. Conclusions Hazelnut shell and Elbistan lignite, which are important energy sources in Turkey, have important differences in their physical, chemical, and thermal properties. Blending these two fuel samples may have the potential for compensating the negative aspects of each other. However, the thermal reactivity of each sample during combustion is highly distinctive. The addition of hazelnut shell into Elbistan lignite affects the combustion system in such a way that volatilization and gaseous phase combustion become more important compared to the combustion of coal alone. Burnouts obtained for the blend samples were not always at the values which are calculated from the fractions of the experimental values of the parent samples. The addition of
2083
hazelnut shell up to 8 wt% into Elbistan lignite provided higher burnouts than the expected theoretical ones. However, the addition of 10 wt% biomass played a reducing role on the burnouts. Burnout characteristics of the blends up to 10 wt% during co-combustion can be concluded to show synergistic effect, whereas additive effects are more favorable in the case of the blend having a biomass content of 20 wt%. On the other hand, apparent activation energy, maximum burning rate, and temperature of the maximum burning rate can be concluded to follow the additive behavior for the blend samples. References Amand, L.E., Leckner, B., Eskilsson, D., Tullin, C., 2006. Deposits on heat transfer tubes during co-combustion of biofuels and sewage sludge. Fuel 85, 1313–1322. Coats, A.W., Redfern, J.P., 1964. Kinetic parameters from thermogravimetric data. Nature 201, 68–69. Demirbas, A., 2002. Fuel characteristics of olive husk and walnut, hazelnut, sunflower, and almond shells. Energy Sources 24, 215– 221. Diaz-Somoano, M., Unterberger, S., Hein, K.R.G., 2006. Prediction of trace element volatility during co-combustion processes. Fuel 85, 1087–1093. Elled, A.L., Amand, L.E., Leckner, B., Anderson, B.A., 2006. Influence of phosphorus on sulphur capture during co-firing of sewage sludge with wood or bark in a fluidized bed. Fuel 85, 1671–1678. Ferrer, E., Aho, M., Silvennoinen, J., Nurminen, R.V., 2005. Fluidized bed combustion of refuse-derived fuel in presence of protective coal ash. Fuel Processing Technology 87, 33–44. Folgueras, M.B., Diaz, R.M., Xiberta, J., Prieto, I., 2003. Thermogravimetric analysis of the co-combustion of coal and sewage sludge. Fuel 82, 2051–2055. Gulyurtlu, I., Boavida, D., Abelha, P., Lopes, M.H., Cabrita, I., 2005. Cocombustion of coal and meat and bone meal. Fuel 84, 2137–2148. Hartmann, D., Kaltschmitt, M., 1999. Electricity generation from solid biomass via co-combustion with coal-energy and emission balances from a German case study. Biomass and Bioenergy 16, 397–406. Haykiri-Acma, H., Yaman, S., Kucukbayrak, S., 2006. Effect of heating rate on the pyrolysis yields of rapeseed. Renewable Energy 31, 803– 810. Karayigit, A.I., Gayer, R.A., Querol, X., Onacak, T., 2000. Contents of major and trace elements in feed coals from Turkish coal-fired power plants. International Journal of Coal Geology 44, 169–184. Klass, D.L., 1998. Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San Diego. Laursen, K., Grace, J.R., 2002. Some implications of co-combustion of biomass and coal in a fluidized bed boiler. Fuel Processing Technology 76, 77–89. Martin, C., Villamanan, M.A., Chamorro, C.R., Otero, J., Cabanillas, A., Segovia, J.J., 2006. Low-grade coal and biomass co-combustion on fluidized bed: exergy analysis. Energy 31, 330–344. Pronobis, M., 2006. The influence of biomass co-combustion on boiler fouling and efficiency. Fuel 85, 474–480. Raghunathan, K., Gullett, B.K., 1996. Role of sulfur in reducing PCDD and PCDF formation. Environmental Science and Technology 30, 1827–1834. Sanchez-Hervas, J.M., Armesto, L., Ruiz-Martinez, E., Otero-Ruiz, J., Pandelova, M., Schramm, K.W., 2005. PCDD/PCDF emissions from co-combustion of coal and PVC in a bubbling fluidized bed boiler. Fuel 84, 2149–2157. Steenari, B.M., Lindqvist, O., 1999. Fly ash characteristics in cocombustion of wood with coal, oil or peat. Fuel 78, 479–488. TAPPI T203 om-88, 1988. Alpha-, beta- and gamma-cellulose in pulp.
2084
H. Haykiri-Acma, S. Yaman / Waste Management 28 (2008) 2077–2084
Theis, M., Skrifvars, B.J., Hupa, M., Tran, H., 2006. Fouling tendency of ash resulting from burning mixtures of biofuels. Part 1: Deposition rates. Fuel 85, 1125–1130. Tillman, D.A., 2000. Biomass cofiring: the technology, the experience, the combustion consequences. Biomass and Bioenergy 19, 365–384. Topuz, A., Topakci, M., Canakci, M., Akinci, I., Ozdemir, F., 2005. Physical and nutritional properties of four orange varieties. Journal of Food Engineering 66, 519–523. Ural, S., 2005. Comparison of fly ash properties from Afsin-Elbistan Coal Basin Turkey. Journal of Hazardous Materials 119, 85–92. van Krevelen, D.W., 1993. Coal: Typology–Physics–Chemistry-Constitution. Elsevier Science Ltd., Amsterdam.
van Soest, P.J., 1963. Use of detergents in the analysis of fibrous feeds. II. A rapid method for the determination of fiber and lignin. Journal of Association of Official Analytical Chemists 46, 830. Xiang-guo, L., Bao-guo, M., Li, X., Zhen-wu, H., Xin-gang, W., 2006. Thermogravimetric analysis of the co-combustion of the blends with high ash and waste tyres. Thermochimica Acta 441, 79–83. Yaman, S., Kucukbayrak, S., 2000. Influence of oxidation in aqueous medium on the structure and properties of lignites. Fuel 79, 777–783. Yaman, S., Karatepe, N., Kucukbayrak, S., 2000. Influence of wet oxidation on the surface area and the porosity of some lignites. Fuel 79, 1017–1022.