Effect of biomass on temperatures of sintering and initial deformation of
Sintering temperatures and the initial deformation temperatures of ashes from Turkish Elbistan lignite,and biomass species such as hazelnut shell and rice husk were investigated up to 1450 ?C by Heating Microscope Technique. Sintering temperatures were found 1300?C, 1269?C, and 1320 ?C for hazelnut shell, rice husk, and lignite, respectively, while the intial deformation while the initial deformation temperatures were>1450?C,1370?C, and>1450?C.Lignite/biomass blends were prepared by adding of biomass into coal in the ratios of 5 or10 wt.%, and then effects of biomass presence on sintering temperature and the initial deformation tem-perature were tested. It was determined that the addition of potassium-rich hazelnut shell reduced the sintering temperatures to 919?C and 730 ?C for the blends of 5 and 10 wt.%, respectively. Also, initial defor-mation temperature dropped to 788 ?C in case of the blend of 10 wt.%. Such a big antagonistic in?uence of hazelnut shell on the thermal behaviour of ash is
attributed to the interaction of potassium from biomass with silicon compounds found in mineral matter of lignite. In addition, concentration of CaO may be another reason for this. On the other hand, the presence of rice husk showed limited effect on the sinter-ing temperature as well the initial deformation temperature.
Co-?ring of coal with biomass is of great interest owing to sev-eral factors such
as controlling of carbon dioxide emissions, mak-ing use of the energy potential of biomass, ef?cient removal of residues, and environmental restrictions on land?ll areas, etc.
However, co-utilization of biomass in the existing combustion systems often leads to serious problems such as deactivation of SCR (selective catalytic reduction) catalysts, slagging, fouling, and clinker formation. These deposit formations are generally attributed to the high alkali content of the biomass material. Some biomass, especially the annual biomass has a high alkali content, which may form low melting point ash during combustion. The low melting ash constituents can induce formation of agglomerates, in addition to deposition and corrosion. In case of fluidized bed system, accumulation of the agglomerates may lead to loss of fluidization (defluidization) and unscheduled shutdown of the plant. Formation of melts, which are responsible for the formation of agglomerates, depends on the type of fuel. The formation of low temperature eutectics is considered as the initiator of agglomerates.
Four different fusion characteristics such as initial deformation temperature,
softening temperature, hemispherical temperature,and fluid temperature are taken into account by ASTM fusion test method to evaluate the fusion characteristics of an ash sample. However, it has been observed in Exxon0s pilot plant gasification process that agglomerates are formed during operation several hundred degrees below the initial deformation temperatures of coal ash tested by ASTM procedure.
Agglomeration of ash may be influenced by various parameters such as the inorganic composition, the particle size and size distribution,the bed temperature and atmosphere under which the operation is conducted. Therefore, the determination of the initial deformation temperature may be a predictor to foresee the extent of the ash related problems.
Accordingly, some fouling or slagging indices are used to predict extent to which a particular fuel will slag or foul upon utiliza- tion. These indices are derived from arious techniques which in- clude the ash fusibility temperature (AFT) test, oxide analysis and ash viscosity measurements. The ash fusibility test has been the most accepted method of assessing whether an ash will foul or slag on the heat transfer surfaces of boilers.
There are techniques through which the melting point of bio- mass ash can be raised to reduce the deposition problems. These are use of additives, use of alternative bed materials in the case of ?uidized bed combustion, and co-?ring with other fuels,
e.g. coals. Additives which can raise the softening temperature of ash are kaolin, ceramic, alumina, calcium oxide, magnesium oxide and dolomite.
Turkey has a great deal of potential for agricultural waste bio-mass species. For an instance, about 70% of world hazelnut production is carried out in this country. So its woody shells present an important source of energy. However, alkali content of its ash is quite high.
On the other hand, approximately 90% of ash from rice husk is consisted of acidic SiO2 so rice husk can be given just a different type of biomass species regarding the alkali/acidic constituents in ash.
Besides, lignitic coals are the most important national primary fossil fuels of Turkey that nearly 8% of annual world lignite con- sumption is performed in Turkey. Among these lignites, Elbistan lignite, which has a low coali?cation degree with an
age of Plio- cene, has the highest deposits and it is used in power stations. Its mineral matter content is 37.5%, and the major constituent in bot- tom ash is CaO . Original lignite has some inorganic phases such as calcite (27.4%), quartz (4.8%),
illite (2.6%), and pyrite (1.6%) .
The aim of this study is to compare the ash-related characteris- tics of hazelnut shell, rice husk, and Elbistan lignite, and to inves- tigate effects of the presence of these biomass materials on the sintering temperature and the initial deformation temperature of lignite ash during co-combustion.
2. Material and methods
Elbistan lignite, hazelnut shell, and rice husk were first air-dried in laboratory for two weeks, representative samples were taken,and then particle sizes were reduced to 60 mesh (-0.25 mm).Proximate analyses and the calorific value measurements were carried out according to the related ASTM standards. Ultimate analyses were conducted by a Eurovector EuroEA3000 model elemental analyzer.Major biomass ingredients such as holocellulosics (holocellulosics+ celluloses), lignin, extractives, and a-cellulose in hazelnut shell and rice husk were determined by the following analytical procedures. Extractive components were determined according to ASTM D1105. The bulks remaining after alcohol–benzene extractions were used to obtain
the holocellulosics by means of NaClO2 extraction procedure. The lignin contents of the samples were found according to the method of van Soest . a-Cellulose contents were determined according to TAPPI T203 om-88 standard. Lignite/biomass blends were prepared by properly mixing of ground (-0.25 mm) Elbistan lignite by hazelnut shell or rice husk with the same particle size. The contributions of biomass in these blends were 5 and 10 wt.%. Ashing of lignite was performed at 750 ?, while
biomass species were ashed at 600? according to ASTM D3174 and ASTM E1755
standards, respectively. Ash of lignite/biomass blends was obtained at 750 ?.
Analyses of ashes were carried out by XRF and XRD techniques. For this purpose, Rigaku Primus II XRF, and Panaltic X0Pert PRO (XRD) were used. SEM images of ashes were obtained using a JoelTM Model JSM-T330 operated at 25 kV and linked with an energy dispersive (EDS) attachment. Heating Microscope Technique was applied in order to investigate the thermal behavior of the ashes. For this purpose, Leitz Heating Microscope model 2004 was used. The pretreatment of the ashes for this procedure includes grinding to 63 lm, moisture fixation at about 4–5 wt.%, and
tabletting to form cubic patterns with dimensions of 2 .22 mm. Cubic patterns were then placed on an alumina plate and heated up to 1450 ?. Variations in the size of
the sample at a heating rate of 40 ?/min were monitored, and evaluated according to
DIN 51730. Burning characteristics of the original fuel samples were investigated by
non-isothermal thermogravimetric analysis method using a Shimadzu TG 41 thermal analyzer with a cylindrical- alumina crucible. Initial weight of the samples was 40 mg, temperature was raised from ambient to 900? by a heating rate of 40 ?/min under
dry air flow of 40 mL/min, and then enough hold time was allowed at this temperature to get the fixed final weights. All the experiments were repeated three times to check the reproducibility of the results, and the mean values were used provided that the deviations were within 5%.
3. Results and discussion
Table 1 shows the results of the proximate and ultimate analyses, and the calorific value for both lignite and biomass samples. Although the lowest fixed carbon belongs to hazelnut shell, it has the highest calorific value. This predicts that volatile matter of this biomass is rich in combustible constituents and it makes important contribution to the calorific value. As to ash contents, hazelnut shell has a different situation that its ash content is considerably lower than the other samples. High mineral matter content in Elbistan lignite led to formation of high content of ash. Although its C content is higher than those of biomass species on dry-ash-free basis, its actual C content on original basis and consequently the calorific value is considerably lower. Major organic ingredients of the biomass species are given in Table 2. Hazelnut shell which has a woody structure is rich in lignin. Whereas, holocellulosics showing high reactivity during combustion or other thermal conversion processes such as pyrolysis and gasification are the most important ingredients of rice husk. Fig. 1 illustrates the thermogravimetric analysis curves through which it is possible to see that they have different burning characteristics. Biomass samples are highly reactive at low temperatures, and accordingly important mass losses from these samples took place even before 300 ?. Most of the biomass
matrix is comprised of oxygen containing polymers of cellulose, holocellulosics, and lignin which are linked together with weak ether bonds with bond energies between 380 and 420 kJ/mol. These bonds are sensitive to heat even at low temperatures, and extensive decomposition took place during the initial stages of the thermal decomposition。
Analysis results of the samples
Major organic ingredients of the biomass samples (%, dry).
Fig. 1. Burning curves of the original fuel samples
processes. In this context, the maximum burning rates (Rmax) weredetermined 12.0 mg/min at 160 ? for hazelnut shell, and 11.2 mg/min at 204 ? for rice husk.
On the other hand, low temperature reactivity of lignite is notas much as hazelnut shell or rice husk. Mass losses occurred upon heating regularly up to 650 ? from the
lignite sample, and then some acceleration happened in the mass loss trend that it may be due to fixed carbon burning phenomenon. Rmax was determined as 5.4 mg/min at 720?. Besides, it can be said as to burning rates at lower region (below 650 ?) that
the burning rates were approximately 1.0 mg/min except the rate measured as 2.00 mg/min at233 ?.
Final weights were determined 1.0, 7.0, and 14.0 mg for hazelnut shell, rice husk, and lignite, respectively. If we consider theash contents in Table 1, the remaining ash from hazelnut shell should have been 1.52 mg. The deviation from this theoretical value is 34.2%, which is at no ignorable level. This exhibits that some of the inorganics became volatile as a result of the increase in the ashing temperature from 600 to 900?. Deviations for other two samples are relatively limited. Results of XRF tests applied to ashes produced according to ASTM D3174 and ASTM E1755 standards for lignite (at 750 ?) and biomasses (at 600 ?) were given in Table 3,
while XRD patterns of these ashes are shown in Fig. 2. Main inorganic phases found from Fig. 2 are given in Table 4.
Ash compositions (%).
According to these data, ash of hazelnut shell is so rich in K2Othat its concentration reaches to 40.34%. Sondreal et al.  reported that CaO is the
dominant constituent in the ashes of woody biomass and K2O follows it in the range of 103 wt.%. This high content of potassium explains the reduce in the weight of remaining ash obtained at 900?, since potassium easily becomes volatile at elevated
temperatures and tends to contribute to fly ash. Potassium is dispersed in biomass, and during combustion it is likely to be volatilized with organic species. CaO and Al2O3 follow K2O. On the other hand, SiO2 content of ash is considerably lower than
those of the other samples.
Ash of rice husk is mostly comprised of SiO2. Although, the second richest compound is K2O, its concentration is only 5.04%.
The most important ingredient in the ash of lignite is CaO (ca.65.6 wt.%). This is in good agreement with the results of Strege et al. who reported a CaO content of 65%. Concentrations of SiO2 and SO3 are comparable and also they accompany to CaO as the other major compounds. Moreover, it is seen that K2O content is as low as 0.34%. Similarly, it is reported in literature that CaO is the most extensive ingredient in North Dakota lignites .
Slagging index is usually calculated based on the ratio of basic oxides to acidic oxides according to following equation: Slagging Index
Slagging indices for hazelnut shell, rice husk, and lignite have been found as 6.52, 0.08, and 4.39, respectively. These values show that the potentials of two biomass species for deposit formation are highly different to each other. On the other hand, although hazelnut shell and lignite gave apparently high values of slagging indices, it is assumed that intensive slagging occurs when the slagging index is remained in the 0.75 range and moving away from this range, in either direction, decreases the slagging intensity.
XRD analysis revealed that silicon is found in the ash of hazelnut shell in both quartz and alumina silicates forms. On the otherhand, orthoclase carries the whole potassium in the ash.Ash of rice husk contains only two different mineral phases. Oneof which, SiO, represents most of the ash. The first three minerals having high 2
intensities in lignite ash belong to calcium containing minerals which followed by sodium and potassium compounds.Besides, SiO is a minor inorganic phase as 2
confirmed by the XRFresults.
SEM micrographs of ashes are seen in Fig. 3. SEM images of hazelnut shell and lignite resemble each other to some extent in terms of their porous and disordered
structures. Fine particles with
only several microns can be seen in case of ash of hazelnut shell. Large SiO2 particles and their obvious reflectance are seen in the SEM image of rice husk ash.
Table 5 shows the temperatures of sintering and initial deformation determined by Heating Microscope Technique.
Sintering temperatures of the original samples, which is defined as a temperature at which a rapid increase in thermal conductivity is observed, vary relatively in a narrow range that rice husk has the lowest (1269?) value while lignite showed the
highest (1320 ?)
Table 4 Inorganic phases in ashes (In the Order of decreasing intensity).
Fig. 2. XRD patterns of ashes.
one. The sintering temperature is an inherent nature of the particlesand is governed by the chemical composition and physicalcharacteristics of the particles.On
the other hand, initial deformation temperatures are higherthan the sintering temperatures, and none of them was lower than1370? Since sintering temperatures
are lower than the initialdeformation temperatures for all the samples used in this study,focusing on the sintering temperature may be an easier predictor to evaluate the agglomeration potential. Lin et al. also reportedthat sintering temperature corresponds 80–90% of the meltingtemperature .There are no reliable
ash fusion characteristics of hazelnut shellin literature, so the results found in this study could not be compared with those of others. However, ash fusion characteristics of rice husk were investigated by various researchers. Mansaray and Ghaly tested the fusion properties of rice husk ash which prepared at 600? . They reported that
initial deformation temperatures vary between 1349–1486?. On the other hand, Yin
et al. reported that in case of rice husk when temperature is more than 850 C, the silica and potassium oxide form a liquid phase that fuse onto the surface of the rice husk char particles, forming a glass-like barrier . Although potassium content of hazelnut shell is very high, sintering could not be observed at low temperatures. The lack of enough silica content to form such a glass-like structure may be the reason for not formation of sintering at low temperatures. InFig. 3. SEM micrographs of ashes.
Temperatures of sintering and initial deformation of ashes
This context, Matsuoka et al. reported that interaction of alkaliand alkaline-earth metals cations with inherent minerals such as quartz and kaolinite reduces the melting point of the original mineral particles. The formed silicates and alumina silicates contribute to agglomeration of ash particles.
Joutsenoja et al. measured the temperatures of the burning coal particles in fluidized bed by pyrometric method, and they reported that average temperature of the coal particles can be considered to be 100–200? higher than the bed temperature
depending on the operation conditions. From this point of view, it can be said that the initial deformation temperatures for any of the original fuel samples used in this study are high enough for utilizing them in fluidized bed combustors operating under usual conditions. This is also in accordance with the calculated slagging indices predicting that they have indices outside of the indicated interval in which severe agglomeration may occur.
Blending the lignite with hazelnut shell so affected the sintering temperature of lignite that it decreased from 1320?to the levels at which serious agglomeration
problems may take place. That is,addition of hazelnut shell reduced the sintering temperature to 919?in case of 5% addition. Moreover, the sintering
temperaturefurther reduced to 730? when lignite was blended with hazelnutshell as
much as 10 wt.%. It is believed that sintering causes interparticle bonds. Sintering is normally defined either as a migrationof holes or lattice vacancies, or as a motion of atoms to a less dense area of the material . The structures of both lignite and biomass are complex, and they are comprised of a number of ingredients. It is possible to suggest that some interactions may take place between ingredients of the different parent materials. Similarly, Bartels et al. determined that involving of limestone to the medium leads a strong increase in the sintering tendency at about 700?. On the
other hand, initial deformation temperature was detected >1450? for the blend of
5%, while it was only 788? for the blend of 10 wt.%. Besides, Kupka et al. reported