Energy Fuels 2010, 24, 4138–4143 Published on Web 07/22/2010
: DOI:10.1021/ef9014313
Ash Deposition Behavior of Upgraded Brown Coal and Bituminous Coal Katsuya Akiyama,*,† Haeyang Pak,† Toshiya Tada,‡ Yasuaki Ueki,§ Ryo Yoshiie,^ and Ichiro Naruse^ †
Mechanical Engineering Research Laboratory, Kobe Steel, Ltd., 1-5-5 Takatsukadai, Nishi-ku, Kobe, 651-2271, Japan, ‡ Coal and Energy Technology Department, Kobe Steel, Ltd., 2-3-1 Niihama, Arai-cho, Takasago, 676-8670, Japan, § Energy Science Division, EcoTopia Science Institute, and ^Department of Mechanical Science and Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan Received November 24, 2009. Revised Manuscript Received June 30, 2010
Ash with a low melting point causes slagging and fouling problems in pulverized coal combustion boilers. Ash deposition on the heat exchanger tubes reduces the overall heat transfer coefficient due to its low thermal conductivity. The ash composition in coal and operational conditions in boilers such as heat load greatly affect the ash deposition behavior. This study focuses on the relationship between ash deposition characteristics and ash melting characteristics. Upgraded brown coal (UBC) and two types of bituminous coal with different melting temperatures and ash compositions were used as samples. The deposition tests were conducted using a refractory furnace. A water-cooled tube was inserted into the furnace to make the ash adhere to it. The molten slag fraction in ash was estimated by means of chemical equilibrium calculations. The results showed that the ash deposition characteristics have a close relationship to the ash melting characteristics. UBC with a low ash melting point shows a relatively high rate of ash deposition. The molten slag fraction in ash obtained by the chemical equilibrium calculation correlates with the deposition fraction of ash obtained by experiments, even under coal blending conditions. Therefore, the molten slag fraction in ash obtained by the chemical equilibrium calculations is one of useful indices to predict the blending method with UBC to reduce the deposition fraction of ash.
the slurry dewatering process.1-3 Akiyama et al. also evaluated the combustion characteristics of UBC.4 However, the ash melting point of Indonesian low-rank coal is relatively low. As a result, slagging and fouling problems may occur in pulverized coal combustion boilers. The ash that is deposited on the heat exchanger tubes reduces the overall heat transfer coefficient due to its low thermal conductivity. Therefore, it is necessary to have an appropriate blend of low-rank and high-rank coal to reduce these ash deposition problems. Ash deposition phenomena are known to be influenced by the coal type (ash constituents, melting temperature, distribution of mineral matter, etc.), reaction atmosphere, particle temperature, surface temperature of the heat exchanger tube, and its surface material, flow dynamics, and so forth. A number of reviews relating to the ash deposition characteristics have already been reported.5 For instance, Raask elucidated the deposit initiation,6 and Walsh et al. and Baxter studied deposit characteristics and growth.7,8 Beer et al. tried to develop theories to model ash behavior.9 Benson et al. summarized the behavior of ash formation and deposition during coal combustion.10 Naruse et al. evaluated the ash
1. Introduction Coal is usually categorized into several types according to its properties. It can also be divided into two major types of high-rank and low-rank coal. Generally speaking, bituminous coal, which is a high-rank coal, is most commonly used in industries such as the thermal power generation industry. On the other hand, demand for low-rank coal such as subbituminous coal, lignite, and brown coal is limited due to its lower calorific value and higher moisture content. Therefore, low-rank coal is usually consumed as fuel only in regions located around its local mine. Recently, there have been demands for a technology that enables low-rank coal such as sub-bituminous coal and lignite to be combusted so that limited fossil resources can be used effectively. Certain types of Indonesian low-rank coal have features such as lower amounts of nitrogen, sulfur and ash. Therefore, if the calorific value of such coal can be increased up to the level of bituminous coal by means of an efficient upgrading treatment, it will be possible to use it as a new fossil fuel. Kobe Steel has developed an upgraded brown coal (UBC) process based on *To whom correspondence should be addressed. E-mail: akiyama.
[email protected]. Phone: þ81 78 992 5634. Fax: þ81 78 993 2056. (1) Sugita, S.; Deguchi, T.; Shigehisa, T.; Katsushima, S. Demonstration of UBC Process in Indonesia, ICCS&T, Okinawa, Japan, Oct. 2005. (2) Sugita, S.; Deguchi, T.; Shigehisa, T. Demonstration of a UBC (Upgrading of Brown Coal) Process in Indonesia. Kobe Steel Eng. Rep. 2006, 56 (2), 23–26. (3) Yamamoto, S.; Sugita, S.; Mito, Y.; Deguchi, T.; Shigehisa, T.; Otaka, Y. Development of Upgraded Brown Coal Process, The Institute for Briquetting and Agglomeration (IBA), Savannah, GA, Oct. 2007. (4) Akiyama, K.; Tada, T. Combustion characteristics of upgraded brown coal (UBC), Proc. 6th Int. Symp. Coal Combust., China, 2007, 527-532. r 2010 American Chemical Society
(5) Frandsen, F. J. Energy Fuels 2009, 23 (7), 3347–3378. (6) Raask, E. Mineral impurities in coal combustion, Hemisphere Publishing Corporation: Washington, 1985, 169-189. (7) Walsh, P. M.; Sayre, A. N.; Loehden, D. O.; Monroe, L. S.; Beer, J. M.; Sarofim, A. F. Prog. Energy Combust. Sci. 1990, 16 (4), 327–345. (8) Baxter, L. L. Fuel Process. Technol. 1998, 56 (1-2), 81–88. (9) Beer, J. M.; Sarofim, A. F.; Barta, L. E. Inorganic transformations and ash deposition during combustion. In Engineering Foundation; Benson, S. A., Ed.; ASME: New York, 1992. (10) Benson, S. A.; Jones, M. L.; Harb, J. N. Ash formation and deposition, Chap. 4. In Fundamentals of coal combustion for clean and efficient use. Smoot, L. D., Ed.; Elsevier Science: New York, 1993, p 299.
4138
pubs.acs.org/EF
Energy Fuels 2010, 24, 4138–4143
: DOI:10.1021/ef9014313
Akiyama et al.
Figure 1. Schematic diagram of UBC process.
deposition characteristics under high-temperature conditions.11 Harb et al. predicted ash behavior using a chemical equilibrium calculation.12 Hansen et al. quantified ash fusibility using differential scanning calorimetry,13 and Ichikawa et al. measured the liquid phase ratio of ash using differential thermal analysis.14 Even with these references, however, precise and quantitative knowledge on the deposition of coal ash with a low-melting temperature during coal combustion is as yet insufficient. Consequently, it is necessary to elucidate the key parameters controlling the ash deposition characteristics and to predict them quantitatively even when a blend of several types of coal are combusted. This study elucidates the relationship between ash deposition characteristics and ash melting characteristics. Indonesian UBC and two types of bituminous coal with different melting temperatures and ash compositions were used for the combustion tests. The ash deposition tests were conducted using a refractory furnace. A water-cooled tube was inserted into the furnace to make the ash adhere to it. The molten slag fraction in ash was calculated with chemical equilibrium calculations.
Figure 2. Schematic diagram of pulverized coal combustion refractory furnace.
without changing its other properties. Therefore, the UBC produced from Indonesian low-rank coal has good properties such as a low nitrogen, sulfur, and ash content. However, the melting point of UBC ash is relatively low as a parent coal.
2. UBC Process A schematic diagram of the UBC process is shown in Figure 1. The UBC process developed by Kobe Steel upgrades lowrank coal to fuel with a high calorific value by slurry dewatering and has no chemical reaction because the process operates at a low temperature and low pressure. Kobe steel started developing this UBC process for Australian brown coal in 1993, and the 3 t/d pilot plant for Indonesian UBC has been operated successfully since 2001. Recently, the 600 t/d demonstration plant has been operated in Satui, South Kalimantan. In the UBC process, low-rank coal is first crushed and mixed with oil, and a small amount of asphalt is added to the slurry. This coal-oil slurry is heated under pressure in order to dewater it since the oil soaks into the pores of the coal, efficiently removing the water there. The UBC produced adsorbs very little moisture, because the asphalt can deactivate the active sites effectively. After the coal is separated from the oil, it is dried and pressed to make transportable briquettes. As stated above, the UBC process effectively removes the moisture in low-rank coal and improves its calorific value
3. Experimental Section Figure 2 shows a schematic diagram of the pulverized coal combustion refractory furnace used in this study. The furnace was 3.65 m long and 0.4 m in inner diameter. It was insulated by refractory material to reduce the heat loss. The pulverized coal metrized by the volumetric feeder was transported to the burner with nitrogen gas and burnt in the furnace with the primary air at ambient temperature and the preheated secondary air at 573 K. City gas was supplied to the furnace in order to raise the particle temperature up to 2000 K. Three types of coal were used as the samples, as shown in Table 1. The fuel ratio in the table denotes the content ratio of fixed carbon to volatile matter in the coal. As seen from the table, the fuel ratio of UBC was the lowest and the ash melting temperature of RT coal was the highest among the three coals. Ash produced from UBC especially had a low SiO2 content and high Fe2O3 content. These types of coal were pulverized with 85% in mass fraction less than 75 μm. Table 2 shows the experimental conditions. The heat load was fixed at a given value of 149 kW for all the tests. The oxygen concentration of the flue gas at the furnace outlet was fixed at a given value of 1.0% on a dry basis. The oxygen concentration at the ash deposition tube inserted was between 1.0% and 2.0% on a dry basis. A deposition probe, made of stainless steel (SUS304), was inserted 1.9 m from the burner tile where the temperature was around 1573 K. Both the temperature and gas compositions at that location are similar to those in the slagging area near the burner in the pulverized coal combustion boiler. The length and outer diameter of the ash deposition tube were 0.2 and 0.0318 m, respectively. The deposition probe mainly consisted of a test piece for deposition and a water-cooled probe. Two thermocouples
(11) Naruse, I.; Kamihashira, D.; Khairil; Miyauchi, Y.; Kato, Y.; Yamashita, T.; Tominaga, H. Fuel 2005, 84 (4), 405–410. (12) Harb, J. N.; Munson, C. L.; Richards, G. H. Energy Fuels 1993, 7 (2), 208–214. (13) Hansen, L. A.; Frandsen, F. J.; Dam-Johansen, K. Ash fusion quantification by means of thermal analysis, An Engineering Foundation Conference on Impact of Mineral Impurities in Solid Fuel Combustion, Hawaii, Nov., 1997. (14) Ichikawa, K.; Oki, Y.; Inumaru, J.; Ashizawa, M. Trans. Japan Soc. Mech. Eng., Ser. B 2001, 67, 144-150.
4139
Energy Fuels 2010, 24, 4138–4143
: DOI:10.1021/ef9014313
Akiyama et al.
Table 1. Properties of Coals Tested coal sample
RT
heating value [MJ/kg] proximate analysis [wt%, dry]
ash volatile matter fixed carbon
fuel ratio [-] total sulfur [wt %, daf] ultimate analysis [wt %, daf]
ash fusion temperature [K] (oxidizing)
ash compositions [wt %]
carbon
PS
UBC
28.89 12.53
27.53 6.23
27.17 1.76
33.54 53.93 1.61 0.54 82.70
43.31 50.46 1.17 0.75 79.64
54.76 43.55 0.80 0.22 71.15
hydrogen nitrogen sulfur oxygen (balance) initial deformation hemisperical fluid SiO2
5.30 1.87 0.50 9.61 1441
5.76 1.83 0.70 12.06 1326
5.39 0.92 0.18 22.35 1452
>1823 >1823 69.80
1614 1687 56.90
1687 1804 40.00
Al2O3 CaO TiO2 Fe2O3 MgO Na2O K2O P2O5 MnO V2O5 SO3
20.73 0.48 1.03 4.95 0.66 0.25 0.98 0.20 0.05 0.05 0.40
23.00 2.19 0.57 11.80 2.27 0.04 0.44 0.34 0.03 0.05 1.27
27.85 3.70 0.56 19.95 1.21 0.14 0.40 0.05 0.27 0.05 3.80
Figure 3. Temperature profiles along central furnace axis.
Figure 4. Photos of ash deposition on the tube surface for three types of coal after 100 min (upper photos: tubes in the furnace; bottom photos: tubes out of the furnace).
Table 2. Experimental Conditions coal sample heating value [kW] combustion stoichiometric ratio [-] oxygen concentration at furnace outlet [%] oxygen concentration around ash deposition tube [%] gas and mean particle temperature around ash deposition tube [K] maximum mean particle temperature in furnace [K] gas velocity [m/s] surface temperature of ash deposition tube [K] ash deposition tube diameter [mm] ash deposition tube length [mm] exposure time for ash deposition [min]
RT, PS, UBC 149 1.05 1.0 1.0-2.0 1534-1545 ∼2000 ∼2.6 773 31.8 200 30, 60, 100
were installed between the test piece and the water-cooled probe. The temperatures measured by those thermocouples were assumed to be the surface temperature of the test piece of deposition. The surface temperature was controlled at 773 K by adjusting the flow rate of cooling water inside the tube. The cross-sectional structure and compositions of ash deposit on the tube was observed, using a scanning electron microscope (SEM) and an energy dispersive X-ray spectroscopy (EDX) apparatus. In the experiments, the thermocouple and gas sampling probe were inserted into the furnace to measure the gas temperature and gas concentrations, respectively.
Figure 5. Cross-sectional structure and compositions of deposit on the tube for three types of coal.
4. Results and Discussion
1545 K. The maximum particle temperature measured with a two-color pyrometer reached 2000 K at a location 0.3 m from the burner tile. Photos of ash deposition on the tube surface for three types of coal after 100 min of exposure are shown in Figure 4. For RT coal, numerous fine gray particulates seemed to
4.1. Ash Deposition Behavior. Figure 3 shows the temperature profiles for three types of coal along the central furnace axis measured with a thermocouple. The temperature profiles are similar. The temperature at the location where the ash deposition tube was inserted was between 1534 and 4140
Energy Fuels 2010, 24, 4138–4143
: DOI:10.1021/ef9014313
Akiyama et al.
Figure 7. Experimental results of deposition fraction of ash under coal blending conditions.
Figure 6. Time dependence of deposition fraction of ash for the three types of coal.
deposit on the surface, and some particulates agglomerated with each other. For PS coal, numerous fine brown particles deposited on the surface, and some of the deposited ash was in a molten state. For UBC, on the other hand, most of the deposited ash was in a molten state, and a very thin layer of molten slag formed on the surface of the tube. Figure 5 shows the cross-sectional structure and compositions of the deposit on the tube for the three types of coal. In those figures, the light gray layer indicates the oxidized thin film of the tube made of stainless steel (SUS304). The gray particles show the ash. For RT coal, there are many spherical ash particles on the layer of the oxidized thin film on the tube. Some ash particles with SiO2 and Al2O3 also locally exist in the layer of the oxidized thin film. For PS coal, there are a few ash particles in the layer of the oxidized thin film. However, for UBC, there are no particles on the oxidized thin film on the tube. Additionally, the ash particles with SiO2 and Al2O3 uniformly exist in the layer of the oxidized thin film. It is thought that the molten slag from the UBC ash reacts with the oxidized thin film and forms a homogeneously decentralized molten slag layer on the tube. Figure 6 shows the deposition fraction of ash (φash) for the three types of coal. The deposition fraction of ash is defined in the following equations: Dash ð1Þ φash ¼ Fash- probe 3 t Fash- probe ¼
Ap Fash- furnace Af
Figure 8. Mechanism of ash deposition on the tube. Table 3. Condition of Chemical Equilibrium Calculation for Three Types of Coal Ash and Their Blends temperature (K) gas composition (%) ash composition (wt %)
1273-2073 1.0 O2 19.0 CO2 80.0 N2 SiO2, Al2O3, CaO, TiO2, Fe2O3, MgO, Na2O, K2O, P2O5, MnO, V2O5, SO3
molten state of ash particles affects the ash deposition on the surface of the tube. From these results, we expect that the ash composition such as SiO2, Al2O3, Fe2O3, and alkali metals and molten state of slag will affect the behavior of ash deposition on the tube. Figure 7 shows the experimental results of the deposition fraction of ash under coal blending conditions. The deposition fraction of ash (φash_t=100) after 100 min corresponding to the mixing ratio of ash is measured in the experiment with blended coal combustion. In the figure, the deposition fraction of ash produced from a mixture of RT coal and PS coal shows a linear relation without any local maximum value. However, the deposition fractions of ash produced from a mixture with UBC show a curve with a local maximum value. In particular, the deposition fraction of ash produced from a mixture of PS coal and UBC indicates the highest value when the mixing ratio of UBC ash is around 60%. From these results, we confirmed that the deposition fraction of ash under coal blending conditions affects the molten state of ash based on the blended ash compositions. 4.2. Chemical Equilibrium Calculation. Figure 8 shows the mechanism of ash deposition on the tube. It is thought that the ash particles tend to adhere more to the tube as the amount of molten particles increase. In this study, we tried to evaluate the molten state of ash with chemical equilibrium calculations using Fact Sage Ver. 5.5 software. Table 3 shows
ð2Þ
In eqs 1 and 2, Dash denotes the deposition mass of ash for a certain exposure time (t). Fash-probe and Fash-furnace mean the mass flow rate of ash in the cross-sectional areas of the probe (Ap) and furnace (Af), respectively. The profile of deposition fraction of ash depends on the coal type, as shown in Figure 6. The initial deposition fractions of ash for three types of coal have high values and become steady after 100 min. Therefore, the exposure time for evaluating the deposition fraction of ash is defined as 100 min. We found that the deposition fractions of ash after 100 min are high in the order of UBC, PS coal, and RT coal. In particular, the deposition fraction of UBC is relatively high, even though the ash content is low. We assume that some factors such as particle size, momentum, and viscosity affect the deposition fractions of ash. However, we confirmed that the ash particle size distributions of three types of coal measured by the particle size analyzer are almost the same. Therefore, we consider that the 4141
Energy Fuels 2010, 24, 4138–4143
: DOI:10.1021/ef9014313
Akiyama et al.
the chemical equilibrium calculations used for the three types of coal ash and their blends. In the calculations, the ash
compositions of the blended coal were assumed to be uniform. The chemical equilibrium calculations were performed in 50 K increments to determine the mass percentage of molten slag in ash for temperatures ranging from 1273 to 2073 K. The gas compositions of atmosphere around the ash deposition probe are also shown in Table 3. Figure 9 shows the calculated molten slag fractions for the three types of coal ash. The molten slag fraction is defined as the content ratio of molten slag to total ash. The molten slag formation for the three types of ash starts between 1273 and 1473 K. The molten slag fraction of UBC ash rapidly increases between 1473 and 1523 K; however, its growth rate decreases at temperatures of 1523 K or higher. Although the molten slag fraction for PS ash also increases over 1473 K, the growth rate does not decrease around 1523 K. For RT ash, the growth rate of the molten slag fraction in ash increases very slowly up to 1623 K. The coal ash compositions and calculated results of the molten slag compositions at 1573 K for the three types of coal ash are shown in Table 4. For the RT ash, all of CaO, Na2O, and K2O were contained in the molten slag, but the major components like SiO2, Al2O3, and Fe2O3 are little contained in the molten slag. However, for the PS and UBC ashes, the major components like SiO2, Al2O3, and Fe2O3 are contained in the molten slag. These calculated molten slag fractions in ash shown in Figure 9 correlate well with the experimental results of the deposition fraction of ash, as shown in Figure 6. According to Table 4, the phase change of the main components in ash like SiO2, Al2O3, and Fe2O3 from solid to slag phase may relate to easiness of the ash deposition. As the result of Table 4 suggests that the stable slag compositions are obtained thermodynamically, the molten slag fraction in ash obtained by the chemical equilibrium calculations is a useful index for predicting the deposition fraction of ash. Figure 10 shows the calculated results of molten slag fraction in ash at a temperature of 1573 K under ash blending conditions. In this figure, the molten slag fraction in ash shows a linear relation with no local maximum value, with a mixture of RT ash and PS ash. However, the molten slag fractions show curves with a local maximum value when there is a mixture with UBC ash. In particular, the molten slag fraction in ash produced from a mixture of PS ash and UBC ash shows the highest value when the mixing ratio of UBC ash is around 60%. These calculation results correlate well with the experimental results, as shown in Figure 7. The calculated results of molten slag compositions at 1573 K for a blend of PS and UBC ash are shown in Table 5. The amount of molten slag such as SiO2, Al2O3, Fe2O3, and FeO changes with the mixing of PS and UBC ash; however, the amount of molten slag such as CaO, TiO2, Na2O, and K2O
Figure 9. Calculated results of molten slag fraction in ash for three types of coal ash.
Figure 10. Calculated molten slag fractions in ash under ash blending conditions. Table 4. Coal Ash Composition and Calculated Results of Molten Slag Composition at 1573 K for Three Types of Coal Ash RT (ash) RT (slag) PS (ash) PS (slag) UBC (ash) UBC (slag) SiO2 Al2O3 CaO TiO2 Fe2O3 FeO MgO Na2O K2O P2O5 MnO V2O5 SO3
69.80 20.73 0.48 1.03 4.95 0.00 0.66 0.25 0.98 0.20 0.05 0.05 0.40
22.48 2.88 0.48 1.03 0.03 0.38 0.08 0.25 0.98 0.00 0.05 0.00 0.00
56.90 23.00 2.19 0.57 11.80 0.00 2.27 0.04 0.44 0.34 0.03 0.05 1.27
35.69 12.90 2.19 0.57 3.07 7.86 2.22 0.04 0.07 0.00 0.03 0.00 0.00
40.00 27.85 3.70 0.56 19.95 0.00 1.21 0.14 0.40 0.05 0.27 0.05 3.80
33.99 13.07 3.70 0.56 2.68 7.41 1.08 0.14 0.08 0.00 0.27 0.00 0.00
Table 5. Calculated Results of Molten Slag Compositions at 1573 K for a Blended Condition of PS and UBC Ash
SiO2 Al2O3 CaO TiO2 Fe2O3 FeO MgO Na2O K2O MnO total
PS 100
PS93 þ UBC7
PS84 þ UBC16
PS70 þ UBC30
PS47 þ UBC53
UBC 100
35.69 12.90 2.19 0.57 3.07 7.86 2.22 0.04 0.07 0.03 64.63
37.06 13.38 2.29 0.57 3.21 8.23 2.20 0.05 0.07 0.05 67.09
38.75 13.98 2.43 0.57 3.41 8.73 2.11 0.05 0.08 0.07 70.17
41.21 14.85 2.64 0.57 3.70 9.49 1.95 0.07 0.08 0.10 74.66
43.48 15.63 2.99 0.57 3.87 10.08 1.70 0.09 0.09 0.15 78.65
33.99 13.07 3.70 0.56 2.68 7.41 1.08 0.14 0.08 0.27 62.98
4142
Energy Fuels 2010, 24, 4138–4143
: DOI:10.1021/ef9014313
Akiyama et al.
deposition fraction in ash rapidly increases if the molten slag fraction in ash becomes over around 60%. As a result, the molten slag fraction in ash obtained by the chemical equilibrium calculation correlates with the deposition fraction of ash obtained by experiments even under coal blending conditions. Therefore, the molten slag fraction in ash obtained by the chemical equilibrium calculations is one of useful indices to predict the blending method with UBC to reduce the deposition fraction of ash. 5. Conclusions An ash deposition test was conducted and chemical equilibrium calculations were made for three types of coal in this study. The following findings were obtained: (i) We confirmed that the ash deposition characteristics have a close relationship with the ash melting characteristics. (ii) The molten slag fraction in ash obtained by the chemical equilibrium calculation correlated with the deposition fraction of ash obtained in experiments even under coal blending conditions. (iii) The molten slag fraction in ash obtained by the chemical equilibrium calculations is one of useful indices to predict the blending method with UBC to reduce the deposition fraction of ash.
Figure 11. Relationship between molten slag fraction in ash and deposition fraction of ash.
does not change very much. Therefore, it is considered that the amount of SiO2, Al2O3, and Fe2O3 in ash affects the amount of molten slag produced from a mixture of PS ash and UBC ash. Figure 11 shows the relationship between the molten slag fraction in ash and the deposition fraction of ash. We clearly confirmed that the deposition fraction of ash increases as the molten slag fraction in ash increases. In particular, the
Acknowledgment. The authors would like to thank JCOAL for providing UBC samples.
4143
