- Email: [email protected]
Study on characteristics of self-desulfurization and self-denitrification in biobriquette combustion
Twenty-Seventh Symposium (International) on Combustion/The Combustion Institute, 1998/pp. 2973–2979
STUDY ON CHARACTERISTICS OF SELF-DESULFURIZATION AND SELFDENITRIFICATION IN BIOBRIQUETTE COMBUSTION ICHIRO NARUSE, HEEJOON KIM, GUOQING LU, JIANWEI YUAN and KAZUTOMO OHTAKE Department of Ecological Engineering Toyohashi University of Technology Tempaku-cho, Toyohashi, 441-8580, Japan
The characteristics of self-desulfurization and self-denitrification in biobriquettes were studied experimentally and numerically in this paper. The biobriquette was produced by mixing coal, biomass, desulfurizer, and/or denitrificater under a high-compression pressure condition. The combustion process of biobriquette appears in two stages, namely the volatile combustion stage and the char combustion stage. It was proved that limestone, wasted scallop shell, and calcium hydroxide have effective self-desulfurization capability in biobriquette combustion and that desulfurization mainly happens in the char combustion stage. Comparatively among the three desulfurizers used, calcium hydroxide has the highest desulfurization capability because of its lower calcination temperature, and scallop shell the second because of the larger porosity after calcination. A desulfurization efficiency as high as about 80% can be reached for some kinds of coals using scallop shell as desulfurizer with Ca/S over 3. A modified shrinking-core model was developed to predict the desulfurization efficiency in the char combustion stage, and an approximate agreement was obtained between the predictions and the experiments. It was also found that pulp black liquid, an industrial waste to roll as binder in the biobriquetting process, has both self-denitrification and self-desulfurization capabilities in biobriquette combustion. A denitrification efficiency of about 40% can be obtained by adding the pulp black liquid into biobriquette with about 15% in mass.
Introduction It has been recognized that coal has the highest potential as a future energy source among fossil fuels in the world. Even though coal is mainly used for power generation in large-scale utility boilers, the amount of coal used in domestic stoves and smallscale industrial boilers is still large, especially in developing countries. In China, for example, about 33% of the coal utilized for the energy supply is burned in small boilers and stoves [1]. The direct burning of coals—especially low-grade coals with high ash, sulfur, and nitrogen contents and low heat values—in small boilers and stoves causes serious pollution of the environment, such as dust, SOx, and NOx emissions. Also in China as an example, it was reported that more than 50% of the SOx emission from coal combustion was contributed by small boilers and stoves [2,3]. Coal briquette had been developed as a technique for coal combustion in small boilers and stoves, which can limit the dust emission to a great extent [3,4]. In recent years, biobriquette has been developed as an improved technique on coal briquette, which is produced from a mixture of coal and biomass under high-compression pressure. Some studies [5,6,9,10] have shown that the addition of biomass, like discarded wood chips, bark, and agricultural wastes, in biobriquette can significantly
improve the characteristics of ignition and burnout of coal and decrease the emission of particulate. An optimum addition fraction of biomass in biobriquette has been recommended to be 10% to 30% in mass by considering both combustibility and shapeability [7,9,10]. Many desulfurizers, such as limestone and calcium hydroxide, have been used in flue gas desulfurization devices and fluidized-bed combustion to decrease SOx emission. In a recent study, Naruse et al. [8] have found that wasted scallop shells have a good desulfurization capability, even better than that of limestone, due to its content of CaO and large porosity after calcination. To decrease NOx emission, on the other hand, many denitrificaters, like alkali metals and some oxides, have been also adopted. However, these desulfurizers and denitrificaters have not ever been used in biobriquette combustion until this study. For small boilers and stoves, it is not realistic to install flue gas desulfurization and denitrifization devices for SOx and NOx emission control. Therefore, the self-desulfurization and self-denitrification in biobriquette combustion becomes necessary and important. In this study, a new kind of biobriquette, with mixture of coal, biomass, and some kinds of desulfurizer and/or denitrificater, has been developed. Three kinds of desulfurizers are used here, namely, limestone, wasted scallop shell, and calcium
2973
2974
COAL AND CHAR COMBUSTION
been observed to have both self-denitrification and self-desulfurization capabilities in biobriquette combustion.
Experimental
Fig. 1. Schematic diagram of experimental apparatus.
hydroxide, and their self-desulfurization characteristics have been experimentally investigated. A shrinking-core reaction model is developed to predict the desulfurization efficiency in biobriquette combustion. The pulp black liquid, an industrial waste to roll as binder in the briquetting process, has
The experimental apparatus, schematically shown in Fig. 1, is composed of an electrically heated batch furnace, temperature controllers, a digital balance, and a flue gas analysis system. In the experiment, the furnace was first heated up to a given temperature (1073 K in this study), and then moved upward to heat the biobriquette inside the reactor. The combustion air was supplied from the bottom of the reactor in a fixed flow rate that provides enough oxygen for complete combustion. The mass loss in biobriquette combustion was continuously measured by the digital balance, which gave out the time history of unburnt mass of biobriquette during combustion. The flue gas was continuously sampled and analyzed, and the time histories of concentrations of CO2, O2, as well as SO2 and NOx were obtained. The properties of the tested coals and biomasses are listed in Table 1. The fuel ratio in the table is defined as the content ratio of fixed carbon to volatile matter. The biomasses used here are the wasted bark chip and cornstalk, and the addition in biobriquette is 20% in mass. The particle diameters of coal and biomass are under 1 mm, those of limestone and wasted scallop shell in the range of 297 to 420 lm, and those of calcium hydroxide under 25 lm. The mixture of coal, biomass, desulfurizer, and/or denitrificater is briquetted under the compression pressure of 245.2 MPa into a cylindrical shape with a
TABLE 1 Properties of tested coals and biomasses. Proximate analysis [mass %, dry basis]
Biomass
Calorific value [kJ/kg]
Ash
VM
FC
Fuel ratio [1]
C
H
N
S
QH
QL
SB MI DT NL BR BJ SV SZ XM DS
2.7 13.7 16.7 17.2 17.6 20.5 16.7 5.6 4.9 15.3
53.6 41.6 38.2 36.6 32.7 19.6 19.4 21.5 21.6 16.1
43.7 44.7 45.2 46.2 49.2 59.9 63.9 72.9 73.5 68.6
0.82 1.08 1.18 1.26 1.48 3.05 3.30 3.37 3.41 4.27
71.8 77.4 82.2 79.2 81.9 83.9 90.2 85.4 85.2 84.8
5.2 5.5 5.5 4.9 5.1 4.2 4.9 4.1 3.9 4.2
1.6 1.3 1.9 1.7 1.8 1.0 1.0 0.9 1.0 1.1
1.3 2.6 0.5 0.3 0.4 1.6 0.3 2.1 1.7 2.1
25400 23700 23600 21300 25700 28300 28300 33500 31300 27500
24200 23500 23300 21100 24700 27400 27400 32700 30400 26700
Bark Cornstalk
2.6 2.1
90.6 88.7
1.5 9.2
0.02 0.10
42.5 40.0
5.5 5.2
0.5 2.1
0.0 0.2
18200 16700
17000 15800
Sample
Coal
Ultimate analysis [mass %, d.a.f.]
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
2975
matters in biobriquette are evolved when the biobriquette is heated up to a temperature and burnt in gas phase outside of the biobriquette. The char combustion does not occur on the surface of the biobriquette until the volatile combustion is almost finished. As oxygen diffuses from the surface toward the inside, the char burns to form a flame sheet that keeps moving toward the inside until the center when char is completely burnt out. The burnt-out time of biobriquette depends on the coal properties. The coal with a higher volatile content, consequently a lower fuel ratio, is burnt out in a shorter time. Self-desulfurization in Biobriquette Combustion Fig. 2. Time histories of unburnt fraction in biobriquette combustion (biomass: bark).
dimension of 16 mm in both diameter and height. A biobriquette pellet weights about 5 g. Results and Discussion Combustion Process of Biobriquette Figure 2 shows the time histories of unburnt mass fraction in biobriquette combustion for different coals. The unburnt mass fraction profile has two parts, the first part with a rapid mass loss, and the second with a slower mass loss. These two parts can be considered to correspond to the two combustion stages, respectively; namely, the volatile combustion stage and the char combustion stage. The volatile
Effects of Ca/S ratio The typical time histories of SO2 concentration in flue gas are shown in Fig. 3, measured in biobriquette combustion with limestone as desulfurizer in different Ca/S ratios. The area under the SO2 concentration profile denotes the total SO2 emission. It can be found that the total SO2 emission is decreased when limestone is added into biobriquette. The definition of desulfurization efficiency is introduced as gSOx 4 1 1 SO2(Ca/S4n)/SO2(Ca/S40)
(1)
where SO2(Ca/S40) is the SO2 emission from the biobriquette combustion without desulfurizer, and SO2(Ca/S4n) that with desulfurizer. Ca/S in the equation represents the molar ratio of calcium to sulfur. The desulfurization efficiency can then be calculated in the preceding cases, and the results are also shown in Fig. 3. It is seen that the desulfurization efficiency increases as Ca/S ratio increasing. A desulfurization
Fig. 3. Time history of SO2 concentration in flue gas in biobriquette combustion and desulfurization efficiency (coal: BJ; biomass: bark; desulfurizer: limestone).
2976
COAL AND CHAR COMBUSTION
from under char combustion where the transport is by diffusion. In addition, once out in the gas phase, the sulfur does not contact the biobriquette again and has no chance for capture. In the actual furnace where the biobriquette pellets are placed in a bed, however, this may be different, because the SO2 may contact another pellet and be captured there. This could mean higher desulfurization efficiency in the actual furnace than in the single pellet experiments.
Fig. 4. Correlation between desulfurization efficiency and fuel ratio (biomass: bark; desulfurizer: limestone; Ca/S 4 3).
Fig. 5. Desulfurization efficiencies for different desulfurizers (coal: BJ; biomass: bark).
efficiency of nearly 80% can be reached at the Ca/S of 5. It reveals that the self-desulfurization can be realized by mixing desulfurizers into biobriquette. It is noted that two different parts appear in the SO2 concentration profiles. Comparing with the combustion process shown in Fig. 2, the two parts can be considered corresponding to the SO2 emissions at the volatile combustion stage and the char combustion stage, respectively. It can be clearly seen that the desulfurization mainly occurs in the char combustion stage. The calculated desulfurization efficiencies show that, in the case of Ca/S 4 5, only about 7% of SO2 emission is captured at the volatile combustion stage, whereas about 70% is captured at the char combustion stage. The low desulfurization efficiency under devolatilization can be considered to be because the desulfurizer is not yet calcinated due to the low temperature. Another explanation may be the short residence time of the sulfur in the biobriquette under devolatilization because the gases are transported by bulk flow. This is different
Effects of coal type As the desulfurization mainly occurs under char combustion, a higher desulfurization efficiency might be reached if more sulfur is held by char. This is verified by the experimental results shown in Fig. 4. Among the eight tested coals, those coals with higher fuel ratios generally have higher desulfurization efficiencies. This is because the sulfur content in char is generally proportional to the fixed carbon content in coal. Effects of desulfurizer type As a kind of typical and abundant sea shell, scallop shell has been found to be an effective desulfurizer in our previous study [8]. In this study, scallop shell was also used for the self-desulfurization in biobriquette combustion. According to the composition analysis, the scallop shell contains about 55% CaO in mass, which is a little more than that in the limestone used (about 53%). For comparison, calcium hydroxide was also taken as a desulfurizer here. The experiments were performed for these desulfurizers at different Ca/S ratios, and the calculated desulfurization efficiencies are shown in Fig. 5, together with those for limestone. It is found that the scallop shell has higher desulfurization efficiencies than limestone when Ca/S is less than 3. Especially at Ca/S 4 1, the efficiency can reach about three times higher than that of limestone. For calcium hydroxide, it keeps the highest efficiency among the tested three desulfurizers. The difference in desulfurization efficiency could be considered resulting from the difference in calcination characteristics. According to the measurements, calcium hydroxide has the lowest calcination temperature of 673 K, and scallop shell and limestone have higher calcination temperatures of 900 K and 1023 K, respectively. The lower calcination temperature of calcium hydroxide makes the SO2 capture effective even under devolatilization, whereas both limestone and scallop shell seem to have difficulty in capturing SO2 under the temperature lower than 900 K. In addition, the difference in desulfurization efficiency between limestone and scallop shell can also result from differences in pore structure after calcination. Larger porosity was measured in calcinated scallop shells by Naruse et al. [8], which makes the SO2 capture more effective. As Ca/ S ratio increases, however, the superiority of pore
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
structure in scallop shell becomes less significant because of the large amount of desulfurizer. As a result, scallop shell has almost the same desulfurization efficiency as limestone when Ca/S is above 3. Prediction of Self-desulfurization Efficiency in Biobriquette Combustion Model description As discussed in the preceding, the biobriquette combustion process appears in two stages, namely, the volatile combustion stage and the char combustion stage. In our previous study [10], a volume reaction model was developed to describe the volatile combustion, and a shrinking-core reaction model to describe the char combustion. Because the desulfurization mainly happens in the char combustion stage, the shrinking-core reaction model is adopted here as a base for the prediction of self-desulfurization efficiency in the char combustion stage. According to the shrinking-core reaction model, the char combustion proceeds from the surface toward the inside, controlled by the oxygen diffusion in both the ash layer inside the biobriquette and the gas boundary layer around the biobriquette. To simplify the simulation, the biobriquette is considered in a spherical shape. It has been shown in our previous study [9] that a cylindrically shaped biobriquette, with the dimension 2R in both diameter and height, has nearly the same combustion process as a spherically shaped biobriquette with the diameter 2R. As the flame sheet moves from the surface toward an unburnt core with a radius of rc, the elapsed char combustion time tc can be calculated as [11] tc 4 R/AC0 [1/3 (1/Kc 1 R/Dc)(1 1 r3c /R3) ` R/2Dc (1 1 rc2/R2)]
(2)
Here, R is the radius of biobriquette, Kc the oxygen diffusion coefficient in gas boundary layer, Dc the effective oxygen diffusion coefficient in ash layer, A the volume of reactable char per mole oxygen, and C0 the oxygen concentration in atmosphere. Dc is strongly related to the porosity in ash layer, which in turn depends on the amount of biomass addition and the briquetting compression pressure, and its calculation formulate has been derived previously [10]. To simulate the desulfurization process under char combustion, it is assumed that (1) sulfur and desulfurizer are uniformly distributed in biobriquette, (2) only C ` O2 → CO2 and S ` O2 → SO2 reactions take place under char combustion, and (3) the diffusion rate of flue gas toward the outside is equal to that of oxygen toward the inside. The formed SO2 at the flame sheet is captured by desulfurizers during the char burning process and the diffusing process through the ash layer. The desulfurization reaction is considered as CaO ` SO2 ` 1/2O2 → CaSO4,
2977
and the conversion rate of CaO is taken from the result of Kojima et al. [12] as dxCaO/dt 4 kSO2 (D 1 xCaO) CSO2
(3)
kSO2 4 403.7 exp(16750 T) (4) Here, xCaO is the converted fraction of CaO to CaSO4, CSO2 the molar fraction of SO2, t the reaction time, and D the maximum conversion fraction of CaO to CaSO4, which takes values of 0.87 for scallop shell and 0.36 for limestone as determined in previous experiments [8]. The captured SO2 during the reaction time Dt can then be calculated by integrating equation 3 as DW 4 qCaO (D 1 xCaOt) [1 1 exp(1kSO2 CSO2 Dt)] (5) where qCaO is the molar density of CaO in char, and xCaOt the converted CaO at time t. xCaOt can be calculated similarly by integrating equation 3, and takes value of 0 at t 4 0 when the desulfurizer first reacts with SO2. To calculate the desulfurization efficiency, a numerical method is used here, in which the spherical biobriquette is divided into a great number (N) of small spherical shells with the same thickness (Dr 4 R/N). The totally captured SO2 should be the sum of those captured during char burning at each shell and those captured during the diffusion process through each shell. According to equation 5, the captured SO2 during char burning at shell i is calculated as DWi 4 qCaO D [1 1 exp(1kSO2 CSO2,i Dti)] DVi (6) where DVi is the volume of shell i, and Dti the char burnout time of shell i. The uncaptured SO2 during the char burning at shell i, if it exists, will diffuse toward the outside and will be captured by the residual desulfurizer in the ash layer. The captured SO2 during the diffusion process through shell j is calculated as DWij 4 qCaO (D 1 xCaO,ij) [1 1 exp(1kSO2 CSO2, ij Dtij)] DVj (7) where DVj is the volume of shell j, Dtij the diffusion time through shell j, xCaO,ij the converted CaO fraction at shell j before the diffusion of this part of SO2 through the shell. By summing these two parts of captured SO2 through all shells, the totally captured SO2 can be calculated, and then the desulfurization efficiency can be obtained. Comparison between predictions and experiments The desulfurization efficiencies were calculated in the cases with scallop shell and limestone as desulfurizers in different Ca/S ratios and are shown in
2978
COAL AND CHAR COMBUSTION
Fig. 6 together with the experimental results for comparison. An agreement can be seen between predictions and experiments for scallop shell. For limestone, there exist some deviations between predictions and experiments, which might result from the determination of the conversion rate and the maximum conversion fraction of CaO to CaSO4. The comparison reveals that the developed model can describe the self-desulfurization in char combustion of biobriquette. Further improvement on the model is being considered by introducing the description of the diffusion-controlled process of sulfur capture inside desulfurizer particles. Fig. 6. Comparison between predictions and experiments (coal: BJ; biomass: bark).
Fig. 7. Time histories of NOx concentration in flue gas with and without addition of pulp black liquid.
Fig. 8. Effect of pulp black liquid content on denitrification efficiency (coal: DS; biomass: cornstalk).
Self-denitrification in Biobriquette Combustion The pulp black liquid, an industrial waste, was rolled as binder in the biobriquetting process. It was observed in the experiments that it has a denitrification capability in biobriquette combustion. Fig. 7 gives the measured time histories of NO concentration in flue gas in the biobriquette combustion with and without addition of pulp black liquid. The area under the NO concentration profile denotes the total NO emission. It is seen that the NO emission in the char combustion stage is decreased by the addition of pulp black liquid. By calculating the denitrification efficiency in the same way as in the case of desulfurization, an efficiency of about 30% can be reached when the addition of pulp black liquid is about 10% in mass. Further increase of pulp black liquid addition can result in a little increase in denitrification efficiency as shown in Fig. 8. With the same amount of pulp black liquid addition, the denitrification efficiencies are different for those coals having different nitrogen contents in char. The coals with higher fuel ratios are found to have higher denitrification efficiencies because the denitrification mainly happens under char combustion. It was also found that the pulp black liquid has not only the self-denitrification capability but also the self-desulfurization capability in biobriquette combustion. Even though the pulp black liquid itself contains sulfur (1.5% in mass as measured), which results in an increase of SOx emission under devolatilization as shown in Fig. 9, the total SOx emission is still decreased. The desulfurization by pulp black liquid happens also under char combustion similar to those desulfurizers already discussed. The desulfurization efficiency reaches about 16.2% when the addition of pulp black liquid is about 16% in mass. When both pulp black liquid and scallop shell are added, as shown also in Fig. 9, SO2 is captured not only under char combustion but also under devolatilization. The total desulfurization efficiency reaches about 64.6%. Although the desulfurization efficiency by pulp black liquid is not as high as those desulfurizers, its capability of desulfurization as well
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
2979
The pulp black liquid has been observed to have both self-denitrification and self-desulfurization capabilities in biobriquette combustion. A denitrification efficiency of about 40% can be obtained by adding the pulp black liquid of over 15% in mass, however. Acknowledgments
Fig. 9. Time histories of SO2 concentration in flue gas with and without addition of pulp black liquid (coal: DS; biomass: cornstalk).
as denitrification is significant for pollutant control in biobriquette combustion. The mechanism of self-denitrification and self-desulfurization by pulp black liquid is not well understood, but the content of sodium hydroxide (NaOH) in pulp black liquid (normally about 30% in mass [13]) might take the function of denitrification and desulfurization. Both NaNO3 and Na2SO4 have been detected in the ash of biobriquette by X-ray diffractometer. Conclusions The SOx and NOx emissions from biobriquette combustion can be controlled by adding desulfurizers and denitrificaters into biobriquette. Limestone, wasted scallop shell, and calcium hydroxide have been proved to have self-desulfurization capability in biobriquette combustion. The desulfurization mainly happens under char combustion, whereas under devolatilization, only calcium hydroxide can capture a little SO2 because of its lower calcination temperature. Comparatively, calcium hydroxide has the highest self-desulfurization capability, and scallop shell the second. When Ca/S is above 3, the desulfurization efficiency can reach as high as 80% by using scallop shell as desulfurizer. A modified shrinking-core model has been developed to predict the desulfurization efficiency under char combustion. An approximate agreement has been obtained between the predictions and the experiments. The developed desulfurization model, together with previously developed combustion models, will provide significant guidance for the production and application of biobriquette.
With deep regret we wish to acknowledge the untimely death of our coauthor, Professor Kazutomo Ohtake, who passed away in an airplane accident on September 26, 1997, in Indonesia. We mourn his death with deep grief and appreciate his guidance and efforts in this study. The partial funding by the Ministry of Education, Japan, is also acknowledged.
REFERENCES 1. Xu, X. C. and Zhang X. Y., in 2nd International Symposium on Coal Combustion, China Science Press, 1991, pp. 8–21. 2. Sugawara, K., Abe, K., Sugawara, T., and Sholes, M. A., J. Jpn. Energy 74:205–212 (1995). 3. Kim, H. J., Hashimoto, S., Ona, S., Matsui, K., and Sadakata, M., J. Jpn. Energy 76:205–213 (1997). 4. Ford, M. W. J., Cooke, M. J., and Gibbs, B. M., Trans Inst. Chem. Eng. 69 (part B):167–172 (1991). 5. Maruyama, T. and Kamide, M., in 3rd International Symposium on Coal Combustion, China Science Press, 1995, pp. 561–567. 6. Maruyama, T., J. Jpn. Energy 74:70–76 (1995). 7. Kamide, M., “Development of Bio-coal Production Using Abroad Low Grade Coal and Biomass as Materials,” Hokkaido Industrial Research Institute, 1990. 8. Naruse, I., Nishimura, K., and Ohtake, K., Kagaku Kogaku Ronbunshu 21:904–909 (1995). 9. Lu, G. Q., Kim, H. J., Naruse, I., Ohtake, K., and Kamide, M., Kagaku Kogaku Ronbunshu 23:404–412 (1997). 10. Lu, G. Q., Kim, H. J., Naruse, I., Ohtake, K., and Kamide, M., Kagaku Kogaku Ronbunshu 23:954–961 (1997). 11. Levenspiel, O., in Chemical Reaction Engineering, Wiley, New York, 1972, p. 362. 12. Kojima, T., Take, K., Kunii, D., and Furusawa, T., J. Chem. Eng. Jpn. 18:432–438 (1985). 13. Chen, R. Y., in Production and Pollutant Control in Paper Industry, China Environmental Science Press, 1991, p 302.
STUDY ON CHARACTERISTICS OF SELF-DESULFURIZATION AND SELFDENITRIFICATION IN BIOBRIQUETTE COMBUSTION ICHIRO NARUSE, HEEJOON KIM, GUOQING LU, JIANWEI YUAN and KAZUTOMO OHTAKE Department of Ecological Engineering Toyohashi University of Technology Tempaku-cho, Toyohashi, 441-8580, Japan
The characteristics of self-desulfurization and self-denitrification in biobriquettes were studied experimentally and numerically in this paper. The biobriquette was produced by mixing coal, biomass, desulfurizer, and/or denitrificater under a high-compression pressure condition. The combustion process of biobriquette appears in two stages, namely the volatile combustion stage and the char combustion stage. It was proved that limestone, wasted scallop shell, and calcium hydroxide have effective self-desulfurization capability in biobriquette combustion and that desulfurization mainly happens in the char combustion stage. Comparatively among the three desulfurizers used, calcium hydroxide has the highest desulfurization capability because of its lower calcination temperature, and scallop shell the second because of the larger porosity after calcination. A desulfurization efficiency as high as about 80% can be reached for some kinds of coals using scallop shell as desulfurizer with Ca/S over 3. A modified shrinking-core model was developed to predict the desulfurization efficiency in the char combustion stage, and an approximate agreement was obtained between the predictions and the experiments. It was also found that pulp black liquid, an industrial waste to roll as binder in the biobriquetting process, has both self-denitrification and self-desulfurization capabilities in biobriquette combustion. A denitrification efficiency of about 40% can be obtained by adding the pulp black liquid into biobriquette with about 15% in mass.
Introduction It has been recognized that coal has the highest potential as a future energy source among fossil fuels in the world. Even though coal is mainly used for power generation in large-scale utility boilers, the amount of coal used in domestic stoves and smallscale industrial boilers is still large, especially in developing countries. In China, for example, about 33% of the coal utilized for the energy supply is burned in small boilers and stoves [1]. The direct burning of coals—especially low-grade coals with high ash, sulfur, and nitrogen contents and low heat values—in small boilers and stoves causes serious pollution of the environment, such as dust, SOx, and NOx emissions. Also in China as an example, it was reported that more than 50% of the SOx emission from coal combustion was contributed by small boilers and stoves [2,3]. Coal briquette had been developed as a technique for coal combustion in small boilers and stoves, which can limit the dust emission to a great extent [3,4]. In recent years, biobriquette has been developed as an improved technique on coal briquette, which is produced from a mixture of coal and biomass under high-compression pressure. Some studies [5,6,9,10] have shown that the addition of biomass, like discarded wood chips, bark, and agricultural wastes, in biobriquette can significantly
improve the characteristics of ignition and burnout of coal and decrease the emission of particulate. An optimum addition fraction of biomass in biobriquette has been recommended to be 10% to 30% in mass by considering both combustibility and shapeability [7,9,10]. Many desulfurizers, such as limestone and calcium hydroxide, have been used in flue gas desulfurization devices and fluidized-bed combustion to decrease SOx emission. In a recent study, Naruse et al. [8] have found that wasted scallop shells have a good desulfurization capability, even better than that of limestone, due to its content of CaO and large porosity after calcination. To decrease NOx emission, on the other hand, many denitrificaters, like alkali metals and some oxides, have been also adopted. However, these desulfurizers and denitrificaters have not ever been used in biobriquette combustion until this study. For small boilers and stoves, it is not realistic to install flue gas desulfurization and denitrifization devices for SOx and NOx emission control. Therefore, the self-desulfurization and self-denitrification in biobriquette combustion becomes necessary and important. In this study, a new kind of biobriquette, with mixture of coal, biomass, and some kinds of desulfurizer and/or denitrificater, has been developed. Three kinds of desulfurizers are used here, namely, limestone, wasted scallop shell, and calcium
2973
2974
COAL AND CHAR COMBUSTION
been observed to have both self-denitrification and self-desulfurization capabilities in biobriquette combustion.
Experimental
Fig. 1. Schematic diagram of experimental apparatus.
hydroxide, and their self-desulfurization characteristics have been experimentally investigated. A shrinking-core reaction model is developed to predict the desulfurization efficiency in biobriquette combustion. The pulp black liquid, an industrial waste to roll as binder in the briquetting process, has
The experimental apparatus, schematically shown in Fig. 1, is composed of an electrically heated batch furnace, temperature controllers, a digital balance, and a flue gas analysis system. In the experiment, the furnace was first heated up to a given temperature (1073 K in this study), and then moved upward to heat the biobriquette inside the reactor. The combustion air was supplied from the bottom of the reactor in a fixed flow rate that provides enough oxygen for complete combustion. The mass loss in biobriquette combustion was continuously measured by the digital balance, which gave out the time history of unburnt mass of biobriquette during combustion. The flue gas was continuously sampled and analyzed, and the time histories of concentrations of CO2, O2, as well as SO2 and NOx were obtained. The properties of the tested coals and biomasses are listed in Table 1. The fuel ratio in the table is defined as the content ratio of fixed carbon to volatile matter. The biomasses used here are the wasted bark chip and cornstalk, and the addition in biobriquette is 20% in mass. The particle diameters of coal and biomass are under 1 mm, those of limestone and wasted scallop shell in the range of 297 to 420 lm, and those of calcium hydroxide under 25 lm. The mixture of coal, biomass, desulfurizer, and/or denitrificater is briquetted under the compression pressure of 245.2 MPa into a cylindrical shape with a
TABLE 1 Properties of tested coals and biomasses. Proximate analysis [mass %, dry basis]
Biomass
Calorific value [kJ/kg]
Ash
VM
FC
Fuel ratio [1]
C
H
N
S
QH
QL
SB MI DT NL BR BJ SV SZ XM DS
2.7 13.7 16.7 17.2 17.6 20.5 16.7 5.6 4.9 15.3
53.6 41.6 38.2 36.6 32.7 19.6 19.4 21.5 21.6 16.1
43.7 44.7 45.2 46.2 49.2 59.9 63.9 72.9 73.5 68.6
0.82 1.08 1.18 1.26 1.48 3.05 3.30 3.37 3.41 4.27
71.8 77.4 82.2 79.2 81.9 83.9 90.2 85.4 85.2 84.8
5.2 5.5 5.5 4.9 5.1 4.2 4.9 4.1 3.9 4.2
1.6 1.3 1.9 1.7 1.8 1.0 1.0 0.9 1.0 1.1
1.3 2.6 0.5 0.3 0.4 1.6 0.3 2.1 1.7 2.1
25400 23700 23600 21300 25700 28300 28300 33500 31300 27500
24200 23500 23300 21100 24700 27400 27400 32700 30400 26700
Bark Cornstalk
2.6 2.1
90.6 88.7
1.5 9.2
0.02 0.10
42.5 40.0
5.5 5.2
0.5 2.1
0.0 0.2
18200 16700
17000 15800
Sample
Coal
Ultimate analysis [mass %, d.a.f.]
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
2975
matters in biobriquette are evolved when the biobriquette is heated up to a temperature and burnt in gas phase outside of the biobriquette. The char combustion does not occur on the surface of the biobriquette until the volatile combustion is almost finished. As oxygen diffuses from the surface toward the inside, the char burns to form a flame sheet that keeps moving toward the inside until the center when char is completely burnt out. The burnt-out time of biobriquette depends on the coal properties. The coal with a higher volatile content, consequently a lower fuel ratio, is burnt out in a shorter time. Self-desulfurization in Biobriquette Combustion Fig. 2. Time histories of unburnt fraction in biobriquette combustion (biomass: bark).
dimension of 16 mm in both diameter and height. A biobriquette pellet weights about 5 g. Results and Discussion Combustion Process of Biobriquette Figure 2 shows the time histories of unburnt mass fraction in biobriquette combustion for different coals. The unburnt mass fraction profile has two parts, the first part with a rapid mass loss, and the second with a slower mass loss. These two parts can be considered to correspond to the two combustion stages, respectively; namely, the volatile combustion stage and the char combustion stage. The volatile
Effects of Ca/S ratio The typical time histories of SO2 concentration in flue gas are shown in Fig. 3, measured in biobriquette combustion with limestone as desulfurizer in different Ca/S ratios. The area under the SO2 concentration profile denotes the total SO2 emission. It can be found that the total SO2 emission is decreased when limestone is added into biobriquette. The definition of desulfurization efficiency is introduced as gSOx 4 1 1 SO2(Ca/S4n)/SO2(Ca/S40)
(1)
where SO2(Ca/S40) is the SO2 emission from the biobriquette combustion without desulfurizer, and SO2(Ca/S4n) that with desulfurizer. Ca/S in the equation represents the molar ratio of calcium to sulfur. The desulfurization efficiency can then be calculated in the preceding cases, and the results are also shown in Fig. 3. It is seen that the desulfurization efficiency increases as Ca/S ratio increasing. A desulfurization
Fig. 3. Time history of SO2 concentration in flue gas in biobriquette combustion and desulfurization efficiency (coal: BJ; biomass: bark; desulfurizer: limestone).
2976
COAL AND CHAR COMBUSTION
from under char combustion where the transport is by diffusion. In addition, once out in the gas phase, the sulfur does not contact the biobriquette again and has no chance for capture. In the actual furnace where the biobriquette pellets are placed in a bed, however, this may be different, because the SO2 may contact another pellet and be captured there. This could mean higher desulfurization efficiency in the actual furnace than in the single pellet experiments.
Fig. 4. Correlation between desulfurization efficiency and fuel ratio (biomass: bark; desulfurizer: limestone; Ca/S 4 3).
Fig. 5. Desulfurization efficiencies for different desulfurizers (coal: BJ; biomass: bark).
efficiency of nearly 80% can be reached at the Ca/S of 5. It reveals that the self-desulfurization can be realized by mixing desulfurizers into biobriquette. It is noted that two different parts appear in the SO2 concentration profiles. Comparing with the combustion process shown in Fig. 2, the two parts can be considered corresponding to the SO2 emissions at the volatile combustion stage and the char combustion stage, respectively. It can be clearly seen that the desulfurization mainly occurs in the char combustion stage. The calculated desulfurization efficiencies show that, in the case of Ca/S 4 5, only about 7% of SO2 emission is captured at the volatile combustion stage, whereas about 70% is captured at the char combustion stage. The low desulfurization efficiency under devolatilization can be considered to be because the desulfurizer is not yet calcinated due to the low temperature. Another explanation may be the short residence time of the sulfur in the biobriquette under devolatilization because the gases are transported by bulk flow. This is different
Effects of coal type As the desulfurization mainly occurs under char combustion, a higher desulfurization efficiency might be reached if more sulfur is held by char. This is verified by the experimental results shown in Fig. 4. Among the eight tested coals, those coals with higher fuel ratios generally have higher desulfurization efficiencies. This is because the sulfur content in char is generally proportional to the fixed carbon content in coal. Effects of desulfurizer type As a kind of typical and abundant sea shell, scallop shell has been found to be an effective desulfurizer in our previous study [8]. In this study, scallop shell was also used for the self-desulfurization in biobriquette combustion. According to the composition analysis, the scallop shell contains about 55% CaO in mass, which is a little more than that in the limestone used (about 53%). For comparison, calcium hydroxide was also taken as a desulfurizer here. The experiments were performed for these desulfurizers at different Ca/S ratios, and the calculated desulfurization efficiencies are shown in Fig. 5, together with those for limestone. It is found that the scallop shell has higher desulfurization efficiencies than limestone when Ca/S is less than 3. Especially at Ca/S 4 1, the efficiency can reach about three times higher than that of limestone. For calcium hydroxide, it keeps the highest efficiency among the tested three desulfurizers. The difference in desulfurization efficiency could be considered resulting from the difference in calcination characteristics. According to the measurements, calcium hydroxide has the lowest calcination temperature of 673 K, and scallop shell and limestone have higher calcination temperatures of 900 K and 1023 K, respectively. The lower calcination temperature of calcium hydroxide makes the SO2 capture effective even under devolatilization, whereas both limestone and scallop shell seem to have difficulty in capturing SO2 under the temperature lower than 900 K. In addition, the difference in desulfurization efficiency between limestone and scallop shell can also result from differences in pore structure after calcination. Larger porosity was measured in calcinated scallop shells by Naruse et al. [8], which makes the SO2 capture more effective. As Ca/ S ratio increases, however, the superiority of pore
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
structure in scallop shell becomes less significant because of the large amount of desulfurizer. As a result, scallop shell has almost the same desulfurization efficiency as limestone when Ca/S is above 3. Prediction of Self-desulfurization Efficiency in Biobriquette Combustion Model description As discussed in the preceding, the biobriquette combustion process appears in two stages, namely, the volatile combustion stage and the char combustion stage. In our previous study [10], a volume reaction model was developed to describe the volatile combustion, and a shrinking-core reaction model to describe the char combustion. Because the desulfurization mainly happens in the char combustion stage, the shrinking-core reaction model is adopted here as a base for the prediction of self-desulfurization efficiency in the char combustion stage. According to the shrinking-core reaction model, the char combustion proceeds from the surface toward the inside, controlled by the oxygen diffusion in both the ash layer inside the biobriquette and the gas boundary layer around the biobriquette. To simplify the simulation, the biobriquette is considered in a spherical shape. It has been shown in our previous study [9] that a cylindrically shaped biobriquette, with the dimension 2R in both diameter and height, has nearly the same combustion process as a spherically shaped biobriquette with the diameter 2R. As the flame sheet moves from the surface toward an unburnt core with a radius of rc, the elapsed char combustion time tc can be calculated as [11] tc 4 R/AC0 [1/3 (1/Kc 1 R/Dc)(1 1 r3c /R3) ` R/2Dc (1 1 rc2/R2)]
(2)
Here, R is the radius of biobriquette, Kc the oxygen diffusion coefficient in gas boundary layer, Dc the effective oxygen diffusion coefficient in ash layer, A the volume of reactable char per mole oxygen, and C0 the oxygen concentration in atmosphere. Dc is strongly related to the porosity in ash layer, which in turn depends on the amount of biomass addition and the briquetting compression pressure, and its calculation formulate has been derived previously [10]. To simulate the desulfurization process under char combustion, it is assumed that (1) sulfur and desulfurizer are uniformly distributed in biobriquette, (2) only C ` O2 → CO2 and S ` O2 → SO2 reactions take place under char combustion, and (3) the diffusion rate of flue gas toward the outside is equal to that of oxygen toward the inside. The formed SO2 at the flame sheet is captured by desulfurizers during the char burning process and the diffusing process through the ash layer. The desulfurization reaction is considered as CaO ` SO2 ` 1/2O2 → CaSO4,
2977
and the conversion rate of CaO is taken from the result of Kojima et al. [12] as dxCaO/dt 4 kSO2 (D 1 xCaO) CSO2
(3)
kSO2 4 403.7 exp(16750 T) (4) Here, xCaO is the converted fraction of CaO to CaSO4, CSO2 the molar fraction of SO2, t the reaction time, and D the maximum conversion fraction of CaO to CaSO4, which takes values of 0.87 for scallop shell and 0.36 for limestone as determined in previous experiments [8]. The captured SO2 during the reaction time Dt can then be calculated by integrating equation 3 as DW 4 qCaO (D 1 xCaOt) [1 1 exp(1kSO2 CSO2 Dt)] (5) where qCaO is the molar density of CaO in char, and xCaOt the converted CaO at time t. xCaOt can be calculated similarly by integrating equation 3, and takes value of 0 at t 4 0 when the desulfurizer first reacts with SO2. To calculate the desulfurization efficiency, a numerical method is used here, in which the spherical biobriquette is divided into a great number (N) of small spherical shells with the same thickness (Dr 4 R/N). The totally captured SO2 should be the sum of those captured during char burning at each shell and those captured during the diffusion process through each shell. According to equation 5, the captured SO2 during char burning at shell i is calculated as DWi 4 qCaO D [1 1 exp(1kSO2 CSO2,i Dti)] DVi (6) where DVi is the volume of shell i, and Dti the char burnout time of shell i. The uncaptured SO2 during the char burning at shell i, if it exists, will diffuse toward the outside and will be captured by the residual desulfurizer in the ash layer. The captured SO2 during the diffusion process through shell j is calculated as DWij 4 qCaO (D 1 xCaO,ij) [1 1 exp(1kSO2 CSO2, ij Dtij)] DVj (7) where DVj is the volume of shell j, Dtij the diffusion time through shell j, xCaO,ij the converted CaO fraction at shell j before the diffusion of this part of SO2 through the shell. By summing these two parts of captured SO2 through all shells, the totally captured SO2 can be calculated, and then the desulfurization efficiency can be obtained. Comparison between predictions and experiments The desulfurization efficiencies were calculated in the cases with scallop shell and limestone as desulfurizers in different Ca/S ratios and are shown in
2978
COAL AND CHAR COMBUSTION
Fig. 6 together with the experimental results for comparison. An agreement can be seen between predictions and experiments for scallop shell. For limestone, there exist some deviations between predictions and experiments, which might result from the determination of the conversion rate and the maximum conversion fraction of CaO to CaSO4. The comparison reveals that the developed model can describe the self-desulfurization in char combustion of biobriquette. Further improvement on the model is being considered by introducing the description of the diffusion-controlled process of sulfur capture inside desulfurizer particles. Fig. 6. Comparison between predictions and experiments (coal: BJ; biomass: bark).
Fig. 7. Time histories of NOx concentration in flue gas with and without addition of pulp black liquid.
Fig. 8. Effect of pulp black liquid content on denitrification efficiency (coal: DS; biomass: cornstalk).
Self-denitrification in Biobriquette Combustion The pulp black liquid, an industrial waste, was rolled as binder in the biobriquetting process. It was observed in the experiments that it has a denitrification capability in biobriquette combustion. Fig. 7 gives the measured time histories of NO concentration in flue gas in the biobriquette combustion with and without addition of pulp black liquid. The area under the NO concentration profile denotes the total NO emission. It is seen that the NO emission in the char combustion stage is decreased by the addition of pulp black liquid. By calculating the denitrification efficiency in the same way as in the case of desulfurization, an efficiency of about 30% can be reached when the addition of pulp black liquid is about 10% in mass. Further increase of pulp black liquid addition can result in a little increase in denitrification efficiency as shown in Fig. 8. With the same amount of pulp black liquid addition, the denitrification efficiencies are different for those coals having different nitrogen contents in char. The coals with higher fuel ratios are found to have higher denitrification efficiencies because the denitrification mainly happens under char combustion. It was also found that the pulp black liquid has not only the self-denitrification capability but also the self-desulfurization capability in biobriquette combustion. Even though the pulp black liquid itself contains sulfur (1.5% in mass as measured), which results in an increase of SOx emission under devolatilization as shown in Fig. 9, the total SOx emission is still decreased. The desulfurization by pulp black liquid happens also under char combustion similar to those desulfurizers already discussed. The desulfurization efficiency reaches about 16.2% when the addition of pulp black liquid is about 16% in mass. When both pulp black liquid and scallop shell are added, as shown also in Fig. 9, SO2 is captured not only under char combustion but also under devolatilization. The total desulfurization efficiency reaches about 64.6%. Although the desulfurization efficiency by pulp black liquid is not as high as those desulfurizers, its capability of desulfurization as well
CHARACTERISTICS OF DESULFURIZATION AND DENITRIFICATION IN BIOBRIQUETTES
2979
The pulp black liquid has been observed to have both self-denitrification and self-desulfurization capabilities in biobriquette combustion. A denitrification efficiency of about 40% can be obtained by adding the pulp black liquid of over 15% in mass, however. Acknowledgments
Fig. 9. Time histories of SO2 concentration in flue gas with and without addition of pulp black liquid (coal: DS; biomass: cornstalk).
as denitrification is significant for pollutant control in biobriquette combustion. The mechanism of self-denitrification and self-desulfurization by pulp black liquid is not well understood, but the content of sodium hydroxide (NaOH) in pulp black liquid (normally about 30% in mass [13]) might take the function of denitrification and desulfurization. Both NaNO3 and Na2SO4 have been detected in the ash of biobriquette by X-ray diffractometer. Conclusions The SOx and NOx emissions from biobriquette combustion can be controlled by adding desulfurizers and denitrificaters into biobriquette. Limestone, wasted scallop shell, and calcium hydroxide have been proved to have self-desulfurization capability in biobriquette combustion. The desulfurization mainly happens under char combustion, whereas under devolatilization, only calcium hydroxide can capture a little SO2 because of its lower calcination temperature. Comparatively, calcium hydroxide has the highest self-desulfurization capability, and scallop shell the second. When Ca/S is above 3, the desulfurization efficiency can reach as high as 80% by using scallop shell as desulfurizer. A modified shrinking-core model has been developed to predict the desulfurization efficiency under char combustion. An approximate agreement has been obtained between the predictions and the experiments. The developed desulfurization model, together with previously developed combustion models, will provide significant guidance for the production and application of biobriquette.
With deep regret we wish to acknowledge the untimely death of our coauthor, Professor Kazutomo Ohtake, who passed away in an airplane accident on September 26, 1997, in Indonesia. We mourn his death with deep grief and appreciate his guidance and efforts in this study. The partial funding by the Ministry of Education, Japan, is also acknowledged.
REFERENCES 1. Xu, X. C. and Zhang X. Y., in 2nd International Symposium on Coal Combustion, China Science Press, 1991, pp. 8–21. 2. Sugawara, K., Abe, K., Sugawara, T., and Sholes, M. A., J. Jpn. Energy 74:205–212 (1995). 3. Kim, H. J., Hashimoto, S., Ona, S., Matsui, K., and Sadakata, M., J. Jpn. Energy 76:205–213 (1997). 4. Ford, M. W. J., Cooke, M. J., and Gibbs, B. M., Trans Inst. Chem. Eng. 69 (part B):167–172 (1991). 5. Maruyama, T. and Kamide, M., in 3rd International Symposium on Coal Combustion, China Science Press, 1995, pp. 561–567. 6. Maruyama, T., J. Jpn. Energy 74:70–76 (1995). 7. Kamide, M., “Development of Bio-coal Production Using Abroad Low Grade Coal and Biomass as Materials,” Hokkaido Industrial Research Institute, 1990. 8. Naruse, I., Nishimura, K., and Ohtake, K., Kagaku Kogaku Ronbunshu 21:904–909 (1995). 9. Lu, G. Q., Kim, H. J., Naruse, I., Ohtake, K., and Kamide, M., Kagaku Kogaku Ronbunshu 23:404–412 (1997). 10. Lu, G. Q., Kim, H. J., Naruse, I., Ohtake, K., and Kamide, M., Kagaku Kogaku Ronbunshu 23:954–961 (1997). 11. Levenspiel, O., in Chemical Reaction Engineering, Wiley, New York, 1972, p. 362. 12. Kojima, T., Take, K., Kunii, D., and Furusawa, T., J. Chem. Eng. Jpn. 18:432–438 (1985). 13. Chen, R. Y., in Production and Pollutant Control in Paper Industry, China Environmental Science Press, 1991, p 302.