Effects of Coal Blends on Formation Mechanisms of Ash Aerosol and Ash Deposits during Air and Oxy-Combustion

Paper from the AFRC 2015 conference titled Effects of Coal Blends on Formation Mechanisms of Ash Aerosol and Ash Deposits during Air and Oxy-Combustion

Many coal-fired power plants now burn coal blends instead of a single raw coal because of required low sulfur attainment levels. Mineral matter for the coal blends are likely to be different from those for their parent coals, and are unlikely to be predictable from simple averaging rules. The problem is important because deposit buildup alters the characteristics of heat transfer and pollutant emissions of the boiler. In this work, experiments were conducted on a 100kW rated pilot-scale down-fired self-sustained combustor, firing an Illinois coal, a Powder River Basin (PRB) coal and a 60% Illinois/40% PRB coal blend. Such a 60%/40% blend had been planned for the FutureGen 2.0 project. Air combustion as well as oxy-coal combustion with recycled flue gas (RFG) were investigated. The intent was not only to test how deposit were formed from coal blend but also to relate the size segregated composition of the ash aerosol to the spatially resolved composition within the deposits. To this end, a Berner low pressure impactor (BLPI), a scanning mobility particle sizer (SMPS), and an aerodynamic particle sizer (APS) were utilized to acquire size segregated ash aerosol samples and to measure particle size distribution (PSD). A novel surface temperature controlled ash deposition probe system was used for fouling deposits collection. The results from air combustion show that PSD's measured by BLPI and SMPS/APS agree well with each other. Combustion of Illinois coal will likely produce more ultra-fine particles compared to PRB coal. However, combustion of Illinois-PRB blended coal could somewhat reduce the formation of these ultra-fine particles. Aerosols from combustion of Illinois coal have higher Si and Al, and corresponding lower Ca, Mg, S and Na compared to those from combustion of PRB coal. The elemental concentrations in aerosols from combustion of blended coal lie between those of the parent coals. Comparing to PRB coal, the inside deposits from combustion of Illinois coal have higher Al, K, Fe and Si, while lower S, Ca, Na and Mg, which is consistent with the trends of ash aerosol composition measurements. This agrees with our previous theory that vaporization mode ash aerosols are the main contributor to build up inside layer deposits, and their composition depends on coal composition. Blended coal increased S retention in ash due to higher alkaline-earth mental (AAEM, especially Ca) concentration in PRB coal. Data from oxy-coal combustion are still being processed and will be included in the full paper.

Effects of Coal Blends on Formation Mechanisms of Ash Aerosol and Ash Deposits during Air and Oxy-combustion Zhonghua Zhan, Andrew R. Fry and Jost O.L. Wendt Department of Chemical Engineering and Institute for Clean and Secure Energy, University of Utah, Salt Lake City, UT 84112, USA Abstract Many coal-fired power plants now burn coal blends instead of a single raw coal because of required low sulfur attainment levels. Mineral matter for the coal blends are likely to be different from those for their parent coals, and are unlikely to be predictable from simple averaging rules. The problem is important because deposit buildup alters the characteristics of heat transfer and pollutant emissions of the boiler. In this work, experiments were conducted on a 100kW rated pilot-scale down-fired self-sustained combustor, firing an Illinois coal, a Powder River Basin (PRB) coal and a 60% Illinois/40% PRB coal blend. Such a 60%/40% blend had been planned for the FutureGen 2.0 project. Air combustion as well as oxy-coal combustion with recycled flue gas (RFG) were investigated. The intent was not only to test how deposit were formed from coal blend but also to relate the size segregated composition of the ash aerosol to the spatially resolved composition within the deposits. To this end, a Berner low pressure impactor (BLPI), a scanning mobility particle sizer (SMPS), and an aerodynamic particle sizer (APS) were utilized to acquire size segregated ash aerosol samples and to measure particle size distribution (PSD). A novel surface temperature controlled ash deposition probe system was used for fouling deposits collection. The results from air combustion show that PSD's measured by BLPI and SMPS/APS agree well with each other. Combustion of Illinois coal will likely produce more ultra-fine particles compared to PRB coal. However, combustion of Illinois-PRB blended coal could somewhat reduce the formation of these ultra-fine particles. Aerosols from combustion of Illinois coal have higher Si and Al, and corresponding lower Ca, Mg, S and Na compared to those from combustion of PRB coal. The elemental concentrations in aerosols from combustion of blended coal lie between those of the parent coals. Comparing to PRB coal, the inside deposits from combustion of Illinois coal have higher Al, K, Fe and Si, while lower S, Ca, Na and Mg, which is consistent with the trends of ash aerosol composition measurements. This agrees with our previous theory that vaporization mode ash aerosols are the main contributor to build up inside layer deposits, and their composition depends on coal composition. Blended coal increased S retention in ash due to higher alkaline-earth mental (AAEM, especially Ca) concentration in PRB coal. Data from oxy-coal combustion are still being processed and will be included in the full paper. Keywords: ash, aerosols, deposition, blends 1. Introduction Ash formation is an important concern in coal combustion, due to subsequent formation of slagging deposits, fouling deposits and particulate matter emissions [13]. Slagging is the deposition process for ash that sticks on the furnace walls and heat transfer surfaces within the radiation zone, while fouling deposition is the deposition process for ash that sticks on the steam tubes located in the convection zone. These ash deposition processes (slagging and fouling) on heat transfer surfaces produce resistances to heat transfer, and cause corrosion and accidents [2]. Particulate matter emissions consist of the fine ash escaping from the ash removal system, and consist of a hazardous air pollutants [3]. Therefore, knowledge of ash formation during combustion is essential to help improve existing coal combustion technology as well as to help evaluate oxy-firing technology prior to retrofit from air firing. During the process of ash formation, an important feature is that submicron particles are formed by vaporization and subsequent condensation of semi-volatile metals, while super micron sized particles are generally formed by fragmentation * Corresponding author. Tel.: [phone] E-mail address: [e-mail]. [3]. Hence one describes ash aerosol size distributions in terms of vaporization modes, caused by sequential vaporization and nucleation/condensation of metals, and fragmentation modes caused by burnout of hollow or solid char particles. Little is available in the literature, in which both the ash aerosol is characterized in detail, and in which that information is linked to the spatially resolved composition of the deposit and deposition rates. It was the objective of this work to provide these links and to use them to determine any special features in mechanisms governing the formation of deposit layers during oxy-coal combustion compared to those formed during combustion in air. In addition, many coal-fired power plants now burn coal blends instead of a single raw coal because of required low sulfur attainment levels. Mineral matter for the coal blends are likely to be different from those for their parent coals, and are unlikely to be predictable from simple averaging rules. The problem is important because deposit buildup alters the characteristics of heat transfer and pollutant emissions of the boiler [4-6]. 2. Materials and Methods 2.1 Oxy-fuel combustor (OFC) Ash deposition experiments were conducted in a 100 kW rated down-filed pulverized coal combustion furnace called the oxy-fuel combustor (OFC). This unit is sufficiently small to allow systematic control of inlet gas flow rates and wall temperatures, and yet large enough to simulate the selfsustaining combustion conditions of full scale units, especially in terms of temperatures, coal particle concentrations and mixing [7]. As shown in Fig. 1, the OFC is down-fired and designed with three different zones: an ignition zone (0.61 m ID), a radiation zone (0.27 m ID) and a convection zone (0.15 m × 0.15 m inside square dimension). The ignition zone contains 3 thermocouples to monitor and control the furnace wall temperatures. The convection zone is installed with eight heat exchangers to cool down the flue gas. Preheaters are used for heating up both primary and secondary inlet gases to 480 K. Nine pairs of ports are positioned along the vertical section of the OFC for sampling and observation. Although the flame in the ignition zone is a pulverized coal turbulent diffusion flame, the flow laminarizes downstream, so that by the time the flow reaches the deposit probe positions it is laminar. This has important implications in collecting and segregating "outer" and "inner" deposits on the deposit probe, as described below [8]. Additional details about OFC can be found elsewhere [7]. A bag house system [9], installed after convection zone, was used for fly ash removal, as shown in Fig. 1. A condenser, as shown in Fig. 1, was used to remove the moisture in the flue gas [9]. The condenser is designed to be a two-stage direct contact heat exchanger, with water as the cooling medium. When the cooling water contains dissolved (and suspended) lime (CaO), the condenser acts as both condenser and scrubber to remove both H2O and SO2 from the flue gas. The condenser is located downstream of the bag house. Fig. 1. Schematic of the OFC with recycled flue gas (RFG) loop. 2.2 Ash aerosols and ash deposits sampling systems For detailed ash aerosol information, two groups of instruments were utilized: a Berner low pressure impactor (BLPI) and a scanning mobility particle sizer (SMPS) coupled with an aerodynamic particle sizer (APS, referred to as SMPS/APS). The working principle for BLPI is inertial impaction and size segregated samples of mineral matter are retained on cellulose acetate membranes for later gravimetric and chemical analysis. The BLPI has a 50% aerodynamic cutoff diameters of 15.7, 7.33, 3.77, 1.98, 0.973, 0.535, 0.337, 0.168, 0.0926, 0.0636, and 0.0324 μm. As shown in Fig. 2, the BLPI sampling system includes a water-cooled, nitrogenquenched isokinetic sampling probe, a cyclone, a BLPI, a vacuum pump, a nitrogen gas cylinder, and a flow meter. The working principle of the SMPS/APS is electron mobility and light scattering. The SMPS produces PSD's in real time, in the size ranges of 0.0143 μm to 0.6732 μm, with the APS measuring in the size ranges of 0.532 μm to 20 μm. The SMPS and APS data are combined and shown in the same figure. The SMPS/APS sampling system includes the same water-cooled, nitrogen-quenched isokinetic sampling probe, a nitrogen gas cylinder, a flow meter, clean compressed air, an orifice, SMPS, APS, and a dilution manifold. The SMPS/APS sampling system, with a total dilution ratio of about 200 to 400 to 1, uses two-stage dilutions to prevent further interactions between particles. The SMPS/APS configuration is also presented in Fig. 2. Sampling was as isokinetic as possible for both air and oxy cases, based on calculated furnace flue gas rates and measured sampling flow rates, which were adjusted for air and oxy conditions through changes in the dilution N2 flow rate. A novel temperature controlled deposits collection probe was used for these experiments. A schematic of the probe is included as Fig. 2. The deposition coupon surface temperature was measured using a K type thermocouples installed in a hole in the deposition coupon. The temperature was controlled for a fixed surface temperature in the range 473 K 973 K with a fluctuation of <10 K. Although the probe was inserted into the furnace when it was cold, it always reached the set temperature within 5 minutes, and usually within 2 minutes, which is a short time compared to a typical holding time of ~4 to 7 hours. Clearly, the temperature difference between the surface exposed to the flue gas and that reported here as "skin temperature" would increase and the temperature gradient between the flue gas and the outer deposit surface would decrease, as the surface deposit layer grows. Compressed air is used as the cooling medium that enters at one end of the inner probe, flows from the small holes in the manifold into the hot end to the space between the inner pipe and outer pipe, to exit as shown. Deposit samples were obtained individually from two layers on the horizontal surface (inside and outside) and from the sides of the coupon and from the vertical surface of the end cap on the collection probe, as highlighted in Fig. 3. The horizontal surface is perpendicular to the flow, thus allowing particle deposition by impaction and diffusion, while the vertical and side surfaces are parallel to the gas streamlines in laminar flow, thus allowing deposition only by molecular processes. In this way it was hoped that horizontal and vertical surfaces would allow segregation of samples deposited by different mechanisms. Tab. 1. Coal and coal ash compositions. Illinois PRB Blend Illinois Coal analysis 2.3 Combustion conditions The combustion conditions in this research include air combustion, oxy-coal combustion of 27% inlet O2 with 73% recycled flue gas with ash removed, and oxy-coal combustion of 27% inlet O2 with 73% recycled flue gas with ash, moisture and sulphur removed, combusting three coals. The coals are Powder River Basin (PRB) coal, Illinois coal and PRB-Illinois blended coal. Tab. 1 provides analysis information of the coals used in the program, including proximate, ultimate and mineral matter analysis of an Illinois # 6 and a PRB Black Thunder coal. The calculated composition of a 60% Illinois, 40% PRB blend is also provided. Key differences in the coal compositions that are important for the deposit compositions are sulfur, calcium, iron and silicon. Coal ash analysis 4.94 7.63 53.72 59.57 6.22 5.75 0.78 1.06 0.23 1.96 34.11 24.04 23.69 15.26 33.36 34.97 38.01 42.14 9078 10562 Al2O3 (%) CaO (%) Fe2O3 (%) MgO (%) MnO (%) P2O5 (%) K2O (%) SiO2 (%) Na2O (%) SO3 (%) TiO2 (%) 6 20.18 14.78 18.02 3.22 22.19 10.81 16.46 5.2 11.96 0.89 5.17 2.6 0.03 0.01 0.02 0.1 1.07 0.49 2.1 0.35 1.4 51.22 30.46 42.92 1.06 1.94 1.41 2.79 8.83 5.21 0.98 1.3 1.11 6 10 10 (a) 5 5 10 4 10 3 10 2 10 1 10 Illinois_Air Blend_Air PRB_Air 0 10 0.01 (b) 10 dM/dlogDp (g/g_ash) Fig. 3. Inside, outside, vertical and side deposits (see also additional details in Reference [10]). Blend 3. Results and Discussion 3.1 Particle size distribution Ash aerosol particle size distributions (PSD's) are shown in Fig. 4 measured by both BLPI and SMPS/APS, which is ploted in a way to exam the effect of coal types on ash aerosol PSD under air combustion condition. The data from the two different methods agree very well in both distribution and magnitude. The SMPS/APS data seems to contain additional resolution showing two modes for the size fraction above 1 µm. These modes are fine and coarse fragmentation modes previously identified by Linak et al. [11]. The PSD of the blend coal lies between the parent coals in the vaporization modes (< 1 µm). dM/dlogDp (g/g_ash) Fig. 2. Controlled wall temperature fouling deposit probe and ash aerosol sampling equipment. ASH 9.42 (%) C 63.47 (%) H 5.43 (%) N 1.24 (%) S 3.12 (%) O (diff) 17.32 (%) LOD 9.64 (%) V 36.04 (%) FC 44.9 (%) HHV 11552 (BTU/lb) PRB 0.1 1 aerodynamic diameter (m) 10 4 10 3 10 2 10 1 10 Illinois_Air Blend_Air PRB_Air 0 10 0.01 0.1 1 10 aerodynamic diameter (m) Fig. 4. Ash aerosols PSD measured by (a) BLPI and (b) SMPS/APS for Illinois, blend and PRB coal. Fig. 5 shows the ash aerosol PSD measured by BLPI, which is ploted in a way to exam the effect of combustion conditions on ash aerosol PSD for Illinois, Blend and PRB coal. The results show that, for the high S containing coals (Illinois and Blend), cases of oxy-coal combustion with ash removed have significant higher peaks in vaporization mode, compared to those with ash and sulphur removed. This is not seen for PRB coal, which contains much less sulphur. This is because for high S coal S is an important component in the vaporization mode, and removral of S decreased the S condensation in the vaporization mode particle range. However, RFG with just ash removed (S contained) increased S in the flue gas, this then increased the condensation of S in the ultrafine size range. This is supported by Fig. 6, which shows the mass fraction of S in the size segregated ash aerosols for Illinois coal, Blend coal and PRB coal under various combustion conditions. It shows that for Illinois and Blend coal (high S coals) the mass concentration of S accounts for more than 50% in the vaporization mode. However, this is not found for PRB coal (low S coal). This is because condensation of S is limited in low S coal. 6 10 4 10 3 10 2 10 1 0 1 10 30 Al 3 10 50 2 40 10 Blend_Air Blend_Ash Blend_Ash_H2O_S 0 10 0.01 10 0.1 1 10 aerodynamic diameter (m) aerodynamic diameter (m) 20 15 10 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 6 10 5 5 (c) 0 0.01 18 3 10 0 0.01 10 Fe 0.1 1 10 aerodynamic diameter (m) Fig. 5. Ash aerosols PSD measured by BLPI for Illinois coal, Blend coal and PRB coal under combustion conditions of various RFG cleanup options. The combustion conditions include air combustion, oxy-coal combustion with RFG of ash removed and oxy-coal combustion with RFG of ash, H2O and S removed. The above plots are: (a) Illinois coal; (b) Blend coal; (c) PRB coal. 90 100 Illinois_Air Illinois_Ash Illinois_Ash-H2O-S 80 70 20 0 10 0.1 1 4 2 0.01 Mg 10 8 6 4 6 4 2 0 0.01 0.1 1 0 0.01 10 0.1 aerodynamic diameter (m) P 0.01 0.1 1 10 aerodynamic diameter (m) 14 PRB_Air PRB_Ash PRB_Ash-H2O-S (c) 12 10 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR S 80 4 2 60 40 20 0 10 0.01 0.1 1 0 0.01 10 aerodynamic diameter (m) 8 45 0.1 1 10 aerodynamic diameter (m) 5 50 6 1 aerodynamic diameter (m) 100 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 6 20 10 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR Na 8 30 1 aerodynamic diameter (m) 10 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 8 aerodynamic diameter (m) mass fraction (%) 0.1 aerodynamic diameter (m) 40 10 0 10 12 50 0 1 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 2 10 0.1 4 0 mass fraction (%) mass fraction (%) 40 6 -2 0.01 60 60 8 2 Blend_Air Blend_Ash Blend_Ash-H2O-S (b) 10 mass fraction (%) 0.01 12 mass fraction (%) PRB_Air PRB_Ash PRB_Ash_H2O_S 0 (a) 6 mass fraction (%) mass fraction (%) 1 1 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR K 14 2 10 80 0.1 aerodynamic diameter (m) 8 16 10 mass fraction (%) 1 20 10 20 10 -20 0.01 0.1 30 aerodynamic diameter (m) 4 10 mass fraction (%) dM/dlogDp (g/g_ash) 10 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR Ca 25 10 10 0.1 4 1 Illinois_Air Illinois_Ash Illinois_Ash_H2O_S 10 0.01 (b) 5 10 dM/dlogDp (g/g_ash) dM/dlogDp (g/g_ash) 10 mass fraction (%) (a) 5 mass fraction (%) 6 10 The BLPI data were further analyzed in order to determine the variability of chemistry in the entrained aerosols as a function of coal type and blending. These data are represented in Fig. 7, where prediction is predicted composition based on a linear relationship betwen the parent coals (60% of Illinos coal and 40% of PRB coal). Aerosols from the combustion of Illinois coal have higher Si, K and Al and lower Ca, Mg, S and Na compared to that of the PRB coal. The composition of the blended coal generally lies between the parent coals with the exception of Na, but different from the linear relationship prediction. Ti Si 4 40 4 0.1 1 10 aerodynamic diameter (m) Fig. 6. Mass fraction of S (measured by EDS) in the size segregated ash aerosols for Illinois coal, Blend coal and PRB coal under combustion conditions of various RFG cleanup options. The combustion conditions include air combustion, oxy-coal combustion with RFG of ash removed and oxy-coal combustion with RFG of ash, H2O and S removed. The above plots are: (a) Illinois coal; (b) Blend coal; (c) PRB coal. 3.2 Elemental compositions of ash aerosols mass fraction (%) 0.01 mass fraction (%) 35 2 30 25 20 15 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 10 5 0.01 0.1 1 aerodynamic diameter (m) 3 2 1 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR 0 10 0.01 0.1 1 10 aerodynamic diameter (m) Fig. 7. Elemental compositions of the size segregated ash aerosols (measured by EDS). 3.3 Composition of the inside deposits The fouling deposits were separated manually into inside, outside and vertical deposits, as shown in Fig. 3. These 50 50 Illinois_AIR Blend_AIR Prediction_AIR PRB_AIR (a) inside 40 30 20 10 30 20 10 0 0 Na Mg Al Si P S 50 K Ca Ti Na Fe Si P S K Ca Ti Fe 30 20 10 Na Na Mg Al Si P S K Ca Ti Fe Fig. 8. Elemental compositions of the deposits from different locations of the collection probe (inside, vertical and outside) under air combustion conditions (probe surface temperature is 923 K, holding time is 1 hour, and collection port number is Port 6). Fig. 9 shows the elemental compositions of the inside fouling deposits for Illinois coal, Blend coal and PRB coal under different combustion conditions, which is ploted in a way to compare the effect of coals on the chemistry of the inside fouling deposits. The data of ‘Prediction' are calculated with a linear relationship based on the combustion cases of parent coals (60% of Illinois coal and 40% of PRB coal). It could be seen that the elemental compositions of the inside fouling deposits from Blend coal (60% of Illinois coal and 40% of PRB coal) combustion lay between its parent coals, regardless of the combustion conditions. Comparing to the predicted composition of the inside fouling deposits for Blend coal combustion, the data from experiments show a trend of leaning towards that of Illinois coal. It also could be seen that, comparing to PRB coal, the inside deposits from combustion of Illinois coal have higher Al, K, Fe and Si, while lower S, Ca, Na and Mg, which is consistent with the trends of ash aerosol composition measurements (take air combustion as an example, as shown in Fig. 7). This agrees with our previous theory that vaporization mode ash aerosols are the main contribution to build up inside layer deposits [8,10]. However, the S condensed in the ultrafine size range will not contribute to the inside fouling deposits because they are in the gasous phase at the deposition location. In addition, blended coal increased S retention in ash due to higher alkaline-earth mental (AAEM, especially Ca) concentration in PRB coal. Al Si P S K Ca Ti Fe Fig. 10 shows the elemental compositions of the inside fouling deposits for Illinois coal, Blend coal and PRB coal under combustion conditions of air combustion, oxy-coal combustion with RFG of ash removed and oxy-coal combustion with RFG of ash, H2O and S removed, which is ploted in a way to compare the effect of combustion conditions on the chemistry of the inside fouling deposits. It seems that there is no significant effect of various combustion conditions on inside fouling deposits formation. 50 40 50 (a) Illinois_AIR Illinois_Ash Illinois_Ash_H2O_S inside Blend_AIR Blend_Ash Blend_Ash_H2O_S (b) inside 40 30 mass fraction (%) 10 Mg Fig. 9. Elemental compositions of the inside fouling deposits for Illinois coal, Blend coal and PRB coal under different combustion conditions. Data of ‘Prediction' are calculated with a linear relationship based on the combustion cases of parent coal (60% of Illinois coal and 40% of PRB coal). The combustion conditions are: (a) air combustion; (b) oxy-coal combustion with RFG of ash removed; (c) oxy-coal combustion with RFG of ash, H2O and S removed. 20 10 30 20 10 0 0 Na Mg Al Si P S 50 K Ca Ti Na Fe (c) Mg Al Si P S K Ca Ti Fe PRB_AIR PRB_Ash PRB_Ash_H2O_S inside 40 mass fraction (%) 20 0 Al 0 30 mass fraction (%) mass fraction (%) 40 mass fraction (%) inside vertical outside Mg Illinois_Ash_H2O_S Blend_Ash_H2O_S Prediction_Ash_H2O_S PRB_Ash_H2O_S (c) inside 40 50 Illinois_Ash Blend_Ash Prediction_Ash PRB_Ash (b) inside 40 mass fraction (%) mass fraction (%) samples were then analyzed by EDS for elemental compositions. Deposits at different locations (inside, vertical and outside) of the probe shows apparent dissimilar elemental compositions as shown in Fig. 8; that is, inside and vertical deposits are very similar to each other, but quite different from outside deposits. This implies that the inside and vertical deposits share the same deposition mechanism, namely, thermophoresis, as proposed and proved in previous study of PRB coal [8]. This differs from the dominant formation mechanism, inertial impaction, for the outside deposits. 30 20 10 0 Na Mg Al Si P S K Ca Ti Fe Fig. 10. Elemental compositions of the inside fouling deposits for Illinois coal, Blend coal and PRB coal under combustion conditions of various RFG cleanup options. The combustion conditions include air combustion, oxy-coal combustion with RFG of ash removed and oxy-coal combustion with RFG of ash, H2O and S removed. The above plots are: (a) Illinois coal; (b) Blend coal; (c) PRB coal. 4. Conclusions Experiments were conducted on a 100 kW rated pilotscale down-fired self-sustained combustor, firing an Illinois coal, a PRB coal and a 60% Illinois/40% PRB coal blend. Air combustion as well as oxy-coal combustion with recycled flue gas (RFG) were investigated. A Berner low pressure impactor (BLPI), a scanning mobility particle sizer (SMPS), and an aerodynamic particle sizer (APS) were utilized to acquire size segregated ash aerosol samples and to measure particle size distribution (PSD). A novel surface temperature controlled ash deposition probe system was used for fouling deposits collection. The results from air combustion show that PSD's measured by BLPI and SMPS/APS agree well with each other. Combustion of Illinois coal will likely produce more ultrafine particles compared to PRB coal. However, combustion of Illinois-PRB blended coal could somewhat reduce the formation of these ultra-fine particles. Aerosols from combustion of Illinois coal have higher Si and Al, and corresponding lower Ca, Mg, S and Na compared to those from combustion of PRB coal. The elemental concentrations in aerosols from combustion of blended coal lie between those of the parent coals. Comparing to PRB coal, the inside deposits from combustion of Illinois coal have higher Al, K, Fe and Si, while lower S, Ca, Na and Mg, which is consistent with the trends of ash aerosol composition measurements. This agrees with our previous theory that vaporization mode ash aerosols are the main contributor to build up inside layer deposits, and their composition depends on coal composition. For high S coal S is an important component in the vaporization mode, and removral of S decreased the S condensation in the vaporization mode particle range. Blended coal increased S retention in ash due to higher alkaline-earth mental (AAEM, especially Ca) concentration in PRB coal. In addition, no significant effect of various combustion conditions on inside fouling deposits formation. Deposits of the blend do not obey simple averaging rules for the two components of the blend, therefore, in order to understand mechanisms of deposit formation, one must have access to both the size segregated composition of the ash aerosol and the spatially resolved composition of the deposits. Acknowledgment The authors would like to acknowledge the support from the Illinois Clean Coal Institute and State of Wyoming under the Clean Coal Research Program and National Science Foundation of China (No. 51376061). We also acknowledge Praxair Inc. for contributing the O2 and CO2 supply and thank Chris Coulter, Mbonisi Sibanda and Brian van Devener at the University of Utah for their help in experiments and analysis. References [1] Raask, E. Mineral Impurities in Coal Combustion: Behavior, Problems, and Remedial Measures; Hemisphere Publishing Corporation: New York, 1985. [2] Bryers, R. W. Progress in Energy and Combustion Science 1996, 22, 29-120. [3] Xu, M.; Yu, D.; Yao, H.; Liu, X.; Qiao, Y. Proceedings of the Combustion Institute 2011, 33, 1681-1697. [4] Wang, Q.; Zhang, L.; Sato, A.; Ninomiya, Y.; Yamashita, T. Fuel 2008, 87, 2997-3005. [5] Wang, Q.; Zhang, L.; Sato, A.; Ninomiya, Y.; Yamashita, T. Fuel 2009, 88, 150-157. [6] Zhou, K.; Xu, M.; Yu, D.; Wen, C.; Zhan, Z.; Yao, H. Chin. Sci. Bull. 2010, 55, 3448-3455. [7] Zhang, J. PhD thesis, University of Utah 2010. [8] Zhan, Z.; Fry, A.; Zhang, Y.; Wendt, J. O. L. Proceedings of the Combustion Institute 2014. [9] Morris, W. J. PhD thesis, University of Utah 2011. [10] Zhan, Z.; Fry, A.; Zhang, Y.; Wendt, J. O. L. Proceedings of the Combustion Institute 2015, 35, 2373-2380. [11] Linak, W. P.; Miller, C. A.; Seames, W. S.; Wendt, J. O. L.; Ishinomori, T.; Endo, Y.; Miyamae, S. Proceedings of the Combustion Institute 2002, 29, 441-447.