Environ. Sci. Technol. 2005, 39, 1180-1183
Denitrification Mechanism in Combustion of Biocoal Briquettes HEEJOON KIM* AND TIANJI LI Department of Ecological Engineering, Toyohashi University of Technology, Tempaku-cho, Toyohashi, 441-8580, Japan
Pulp black liquor (PBL), an industrial waste from paper production, has been previously shown to be an effective binder and denitrification agent for coal briquettes. This study investigated the denitrification mechanism of PBL in both the volatile combustion and char combustion stages of coal briquettes. X-ray diffraction and ion chromatography were used to analyze the residual ashes of combustion. The exhaust gas was analyzed by a flue gas analysis system and a Q-mass spectrometry system. The denitrification mechanism of PBL in the volatile combustion stage was found to result from the emission of NH3. The denitrification of PBL in the char combustion stage was associated with the NaOH contained in PBL. The direct reaction of NaOH with NO gas was examined, and some interesting phenomena were observed. Pure carbon or pure NaOH showed only limited reaction with NO. However, the mixture of NaOH and carbon (NaOH + C) significantly enhanced the reaction. This mixture increased the NO removal up to 100%. Subsequently, denitrification lasted for a long time period, with about 25% of NO removal. The pyrolysis characteristic of NaNO3, a compound resulting from denitrification, was also affected by the presence of carbon. In the presence of carbon, the NOx emission resulting from the pyrolysis of NaNO3 was reduced by a factor of 6. Since the denitrification phenomena appeared only in the absence of oxygen, a model of oxygen distribution in a burning coal briquette was employed to explain the reactions occurring in real combustion of coal briquettes.
Introduction In recent years, the energy demand in some developing countries has rapidly increased due to robust economic development. For instance, coal is now a key energy source in China. Unfortunately, flue gas cleaning equipment is not usually installed, even in large-scale thermal power plants, due to financial restraints, and low-grade coals with high sulfur content are widely used in domestic stoves and smallcapacity industrial boilers. Therefore, there is an urgent need for economical methods of desulfurization and denitrification in these countries. Some industrial wastes can be used as effective desulfurizers and denitrificaters. In our previous studies (1-6), pulp black liquor (PBL) was added to biocoal briquettes, and its self-desulfurization and self-denitrification characteristics were investigated. It was found that PBL captures approximately 35% of SOx and 40% of NOx during the combustion process of biocoal briquettes. Analysis of PBL showed that it contains up to 20% NaOH. NaOH has been assumed to play an important role in denitrification; however, the actual mechanism of denitrification of coal briquettes has * Corresponding author e-mail:
[email protected]. 1180
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005
FIGURE 1. Schematic diagram of the vertically integrated experimental apparatus. not been studied in detail. The reactions of sodium compounds at high temperature have been studied by some researchers, but the mechanisms are not clearly understood (7-13). In this study, the denitrification mechanisms of PBL and pure NaOH in volatile combustion and char combustion were examined in parallel. Pyrolysis of coal and PBL was conducted to investigate the components involved in the denitrification mechanism in volatile combustion. Furthermore, direct reaction of NaOH with NO gas was studied. The reaction of NaOH and the reaction of the mixture of NaOH and carbon (NaOH + C) were compared by analyzing the exhaust gas and the residue ash through a gas analysis system and X-ray diffraction (XRD). The pyrolysis characteristics of NaNO3 as the product of denitrification of NaOH were observed. A model of oxygen distribution in a burning coal briquette was employed to explain the reaction of NaOH with NO in the combustion process of coal briquettes.
Experimental Section Two separate experimental apparatuses were used in this study. The combustion experiments of coal briquettes were performed in a vertically integrated electrical furnace, whereas a horizontally integrated electrical furnace was used to investigate the reaction of NaOH + C with NO. The vertical electrical furnace is schematically shown in Figure 1. It consists of an electrically heated furnace with temperature controllers, a digital balance, and a flue gas analyzing system. A packed bed of alumina balls is located at the bottom of the 10.1021/es035358k CCC: $30.25
2005 American Chemical Society Published on Web 01/15/2005
FIGURE 2. Schematic diagram of the horizontally integrated experimental apparatus.
FIGURE 3. Denitrification efficiency of NaOH and pulp black liquor.
furnace to preheat incoming air. The reaction tube and the electrical furnace are integrated into a unit that can be moved up and down vertically. The sample is placed in a basket, which is linked with the upper digital balance, and is positioned in the center of the furnace axis. The changes of mass and SO2, NOx, CO, CO2, and O2 concentrations as a function of time were continuously measured and recorded by the digital balance and the flue gas analyzing system during the combustion process. Samples were made from coal and NaOH with molar ratios of Na to N of 0, 1, 3, and 5. The coal and NaOH were ground to a particle diameter range of under 1 mm. The weight of each sample was 5 g. The electrical furnace was preheated to a determined temperature and then moved upward to ignite the sample. The residual ash was analyzed by using XRD (RINT2500, Rigaku Corp.) and ion chromatography (DX-120, Dionex Corp.). The reaction of NaOH + C with NO was performed in the horizontal furnace consisting of a quartz reactor, an electrical heater, and a temperature controller shown in Figure 2. During the experiment, the sample was set in the center of the quartz reactor and the temperature was kept constant. The flow rate of bulk gas into the reactor was 2 dm3/min. NaOH (96% purity) was mixed with carbon powder (99% purity and 5 µm size). The mixture was put into a ceramic vessel, which was then set in the quartz reactor. The exhaust gas was analyzed continuously using both a NOx analyzer and a Q-mass spectrometry system to monitor the change in concentration of various components. A mixture gas consisting of NO and N2 or NO and Ar was used as the bulk gas for different experiments.
Results and Discussion Our previous results showed that the biocoal briquette with added PBL has self-denitrification characteristics and that NaOH is one of the constituents of PBL. In this study, the denitrification mechanisms of NaOH and PBL were investigated separately. Samples with various molar ratios of Na to N were made by mixing different amounts of NaOH or PBL with coal. The denitrification efficiencies of these samples at 800 °C are shown in Figure 3. The denitrification efficiency increased with the Na/N molar ratio and achieved a maximum value of 45% for both kinds of samples. As shown in Figure 3, PBL had higher denitrification efficiency than NaOH at low Na/N ratios, but had the same maximum efficiency when the Na/N ratio exceeded 2.5. The PBL contains about 20 wt % NaOH, and its main component is lignin (1-3). These results suggest that the NaOH within PBL is responsible for the denitrification capability of PBL and that PBL has higher denitrification efficiency than NaOH alone.
FIGURE 4. (a) Effect of Pulp Black Liquor on the emission of NOx. (b) Effect of NaOH on the emission of NOx. As shown in Figure 4, the emission of NOx underwent two different stages corresponding to the two stages of the combustion process. In the first stage, the volatile combustion stage, NOx was emitted rapidly for a short period. In the second stage, the char combustion stage, NOx was emitted slowly and constantly for a long period. The PBL-induced denitrification mechanisms for the two combustion stages were studied separately. The emission of NOx decreased in both the volatile and char combustion stages following the addition of PBL (Figure 4a). This implies that PBL has denitrification capabilities in both combustion stages. However, the NOx emission decreased only in the char combustion stage for the sample with NaOH as denitrificater. As shown in Figure 4b the emission of NOx in the volatile combustion stage is the same for samples with or without added NaOH. VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1181
FIGURE 5. Conversion compounds of Fuel-N during pyrolysis.
FIGURE 6. X-ray diffraction pattern of residual ash. Two kinds of coal were used as samples. In Figure 4a DS coal was used, and in Figure 4b BJ coal was used. DS coal has a higher nitrogen content than BJ coal (5); this explains why the peak values of NOx concentration in the two figures are different. To investigate the denitrification mechanism of PBL in the volatile combustion stage, the pyrolysis of PBL, coal, and biomass was conducted. Nitride compounds and NH3 in the emission gas from the pyrolysis process were analyzed by GC-MS, and the results are shown in Figure 5. Thirty-eight percent of the nitrogen in PBL was converted to NH3 during the pyrolysis, while the conversion rate for coal was only 5%. NH3 emitted during pyrolysis is responsible for NO reduction via the following global denitrification mechanism:
4NH3 + 6NO f 5N2 + 6H2O
9
the denitrification process of NaOH in the combustion of coal briquettes. To demonstrate the denitrification mechanism, the direct reaction of NaOH with NO was performed in a horizontal furnace (shown in Figure 2). After the furnace was preheated to 900 °C, NO gas at a concentration of 600 ppm, with N2 as the background gas, was sent into the reactor. The quartz reactor with the sample was then moved into the furnace to initiate the experiment. When carbon was used as the reactant, the NO concentration at the outlet of the reactor was almost the same as that at the inlet, as indicated by the dashed line in Figure 7. This means that the NO gas reacted only slightly with pure carbon. When NaOH was used as the reactant, the NO concentration initially decreased slightly and then increased to the original concentration, as shown by the dotted line in Figure 7. However, when a mixture of carbon (0.2 g) and NaOH (0.2 g) (NaOH + C) was used as the reactant, a remarkable denitrification reaction was observed, as shown by the solid line in Figure 7. During the initial 25 min of the reaction, almost 100% of the NO was absorbed. Although our research focused on this first denitrification stage, it is interesting to note that the NO concentration then increased to about 450 ppm and remained at that value for a very long time period. This later stage of denitrification was considered to be due to the reaction of carbon with NO. Our results show that carbon promotes the reaction between NaOH and NO. This phenomenon can explain why the NaOH in PBL can react with NOx, resulting in denitrification. This also means that the carbon in coal briquettes promotes the denitrification reaction of NaOH in PBL during the combustion of coal briquettes. The reaction between NaOH and NO might be considered to be carbon
(1)
This mechanism explains why NaOH does not have denitrification capabilities during the volatile combustion stage. To investigate the denitrification mechanism in the char combustion stage, the residual ashes of samples containing NaOH as denitrificater were analyzed by XRD (Figure 6). The peaks that derived from increasing the amount of NaOH in coal briquettes correspond mainly to NaAlSi4. The peaks of NaNO3 or NaNO2 were not clearly detected by XRD. To examine whether NOx was fixed in ash, a quantitative assessment was performed after the residual ash was dissolved and the NO3-, NO2-, and NH4+ ion concentrations in the solution were analyzed by ion chromatography. About 5% of the total denitrified nitrogen remained in the residual ash, suggesting that NOx is not fixed in the residual ash during 1182
FIGURE 7. Time concentration history of NOx for samples of carbon, NaOH, and NaOH + C.
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005
NaOH + 2NO 98 NaNO3 + 1/2H2 + 1/2N2
(2)
However, since NaNO3 decomposes at about 650 °C, it could not exist under our conditions (900 °C). Therefore, the pyrolysis of NaNO3 and that of the mixture of NaNO3 with carbon were compared. Temperature was evenly increased from 0 to 850 °C over a 2 h period. The time concentration history of NOx was measured and recorded during the process. The pyrolysis characteristics are shown in Figure 8. The addition of carbon lowered the decomposition temperature of NaNO3 and substantially decreased the NOx emission from the NaNO3 decomposition. Specifically, the conversion of nitrate ion to NOx in the decomposition process of NaNO3 was about 60%, and the conversion decreased to about 10% when carbon was added. The residue of the reaction of NaOH + C and NO was analyzed by XRD, and
FIGURE 8. Pyrolysis of NaNO3 and NaNO3 + carbon. the results indicated that sodium existed as Na2CO3 in the residue. This reaction can be considered to be
4NaNO3 + 5C f 2Na2CO3 + 2N2 + 3CO2
(3)
Nitrogen was converted to N2 gas and emitted during the decomposition process. This explains why NaOH has the ability to denitrify but only a small amount of nitrogen is fixed in the ash. Reactions 2 and 3 are considered to be the denitrification mechanism of NaOH. However, reaction 3 is not the denitrification mechanism in combustion of coal briquettes. The denitrification mechanism for combustion of coal briquettes will be discussed later in this study. More detailed research will be necessary to explain the rapid denitrification reactions caused through reactions 2 and 3. Such research will widen the possibility of application of NaOH + C in denitrification fields. It is important to point out that this denitrification process was found to take place only in the absence of oxygen. If oxygen were present in the system, the denitrification reactions would be affected significantly. This raises the question of how the denitrification reaction could occur in the combustion process of coal briquettes where oxygen is present. One way to explore this is by analyzing the distribution of oxygen in a burning coal briquette. As described above (Figure 4), the combustion of coal briquettes is divided into the volatile and char combustion stages. Char combustion plays the major role in the combustion of coal briquettes. During char combustion, oxygen diffuses through the gas film to the surface of the briquette, diffuses through the burnout ash layer to the unreacted core, and finally reacts with coal in the burning layer (Figure 9). In this process, there are three control rates: the gas film diffusion rate, the ash layer diffusion rate, and the chemical reaction rate (46). If the char combustion process is controlled by the gas film diffusion and ash layer diffusion rates, the chemical reaction rate will be much faster than the other two rates. The char combustion process appears in the shrinkingcore reaction model (4-6). This process is controlled by oxygen diffusion through the gas film and the ash layers (4-6). In other words, the reaction of oxygen and coal in the burning layer is much faster than the diffusion of oxygen through the gas film and ash layers. Therefore, the concentration of oxygen in the burning layer (unreacted sphere interface) is zero, as shown in Figure 9, and the combustion reaction in the burning layer is under reducing conditions. The absence of oxygen in the burning layer indicates that reactions 2 and 3 are the most likely to occur in the char combustion of coal briquettes. On the other hand, in the residual ash of coal briquettes, NaAlSi4, and not Na2CO3, is the main constituent (Figure 6).
FIGURE 9. Distribution of O2 in a burning coal briquette. That would indicate that reaction 3 does not occur in the real combustion process of coal briquettes, due to the effects of Al2O3 and SiO2 in coal. After reaction 2, Al2O3 and SiO2 can replace carbon and produce NaAlSi4, as demonstrated by our XRD results:
2NaNO3 + Al2O3 + 8SiO2 f 2NaAlSi4 + N2 + 25/2O2
(4)
Therefore, reactions 2 and 4 can be assumed to be the dominant denitrification reactions of NaOH in char combustion of coal briquettes. Thus, we conclude that reactions 2 and 3 are the reaction mechanisms describing the direct reaction of NaOH with NO, but in char combustion of coal briquettes, the denitrification mechanism is described by reactions 2 and 4. On the basis of our models, carbon plays a key role in the reaction system as the catalyst and reactant.
Literature Cited (1) Naruse, I.; Kim, H. J.; Lu, G. Q.; Yuan, J. W.; Ohtake, K. TwentySeventh Symposium (International) on Combustion; Combustion Institute: Pittsburgh, 1998; p 2973. (2) Kim, H. J.; Naruse, I.; Lu, G. Q.; Ohtake, K.; Kamide, M. Kagaku Kogaku Ronbunshu 1998, 24, 779 (in Japanese). (3) Kim, H. J.; Lu, G. Q.; Li, T. J.; Sadakata, M. Environ. Sci. Technol. 2002, 36, 1607. (4) Kim, H. J.; Lu, G. Q.; Naruse, I.; Yuan, J. W.; Ohtake, K. J. Energy Resour. Technol. 2001, 123, 27. (5) Lu, G. Q.; Kim, H. J.; Yuan, J. W.; Naruse, I.; Ohtake, K.; Kamide, M. Energy Fuels 1998, 12, 689. (6) Levenspiel, O. Chemical Reaction Engineering, 2nd ed.; Wiley: New York, 1972; p 357. (7) McEwan, M. J.; Phillips, L. F. Trans. Faraday Soc. 1965, 62, 1717. (8) Carabetta, R.; Kaskan, W. E. J. Phys. Chem. 1968, 72, 2483. (9) Kaskan, W. E. Tenth Symposium (International) on Combustion; Combustion Institute: Pittsburgh, 1971; p 41. (10) Hynes, A. J.; Steinberg, M.; Schofield, K. J. Chem. Phys. 1984, 80, 2585. (11) Srinivasachar, S.; Helble, J. J.; Ham, D. O.; Domazetis, G. Prog. Energy Combust. Sci. 1990, 16, 303. (12) Schofield, K.; Steinberg, M. J. Phys. Chem. 1992, 96, 715. (13) Vladimir, M.; et al. Twenty-Seventh Symposium (International) on Combustion; Combustion Institute: Pittsburgh, 1998; 1443.
Received for review December 5, 2003. Revised manuscript received November 3, 2004. Accepted November 12, 2004. ES035358K VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1183
