Evaluation of the Performance of Air Dense Medium Fluidized Bed

Aug 13, 2013 - ... Dewatering/Dry Separation of Chinese Lignite in a Vibration Fluidized-Bed Dryer: Effect of Physical Parameters and Operation Condit...
1 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/EF

Evaluation of the Performance of Air Dense Medium Fluidized Bed (ADMFB) for Low-Ash Coal Beneficiation. Part 2: Characteristics of the Beneficiated Coal Ebrahim Azimi,† Shayan Karimipour,‡ Moshfiqur Rahman,‡ Jozef Szymanski,† and Rajender Gupta*,‡ †

Department of Civil and Environmental Engineering, University of Alberta, Edmonton, Alberta T6G 2W2, Canada Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta T6G 2V4, Canada

‡

ABSTRACT: Low-rank coals are widely used as fuel in coal-fired power plants. Continuing the use of these fuels is under huge pressure because of stringent environmental regulations. The air dense medium fluidized bed (ADMFB), which is a dry physical coal beneficiation method, can offer an efficient and economical solution for ash removal. The performance of the ADMFB separator in preparing higher quality coals has been studied by detailed characterization of the ADMFB clean coal products with minimum ash content, maximum organic material recovery, or different feed particle sizes in the present work. The percentage of ash removal for the selected samples with particle size of 1−13.2 mm was between 9 and 22% with respect to the feed sample. An increase of all CHNS components and higher heating value (HHV), regardless of the coal particle size, and a very efficient mercury rejection of between 33.7 and 48.6% were observed for most of the beneficiated samples. Inductively coupled plasma− mass spectrometry (ICP−MS) analysis confirmed the decrease of most of the hazardous elements, such as Pb, Ag, Ba, Cu, Mn, Be, and K, indicating a positive affinity of these elements with the ash-forming minerals of the coal. Some elements, such as As, Se, and Sb, exhibited inconsistent results, which indicates various degrees of organic bonding for these elements. X-ray fluorescence (XRF) analysis of the ashes of the beneficiated coals revealed different decreasing levels of Si, Al, and Mg oxides (main components of clay minerals) and an increase of Na, Ca, and Fe oxides. Lower viscosities, a lower reducing ash fusion temperature (< ∼1250 °C), and consequently, an increased slagging propensity based on a number of simple slagging indices for beneficiated products were obtained. The change in the reactivity of the clean coal products was discussed on the basis of the maximum rate of weight loss (Rmax) and the peak temperature (Tmax) obtained by differential thermogravimetry (DTG). The fine and middle size beneficiated samples showed various degrees of reactivity improvement. Rmax for the middle size was found to increase by at least 84.5%, and Tmax for the same was found to decrease by at least 62 °C.

1. INTRODUCTION Strict environmental regulations and potentially high financial penalties associated with violating the regulatory limits have forced coal power plants to actively reduce their environmental footprint. Combustion of as-mined coal in power plants leads in discharging a large amount of mineral matter in ash form, SOx, and volatile trace elements into the environment. Coal cleaning and beneficiation can play a key role in improving coal quality toward regulation compliance by removing the ash proportion of the run of mine coal prior to combustion and moderate postcombustion issues. Pre-combustion trace element removal is attractive in a sense that it is likely less expensive than postcombustion cleanup and may well prove to be more effective for specific types of elements that potentially can pass through the pollution control devices. Trace elements with a major concern for human health are considered to be As, Cd, Hg, Pb, and Se.1,2 The European Pollutant Emission Register (EPER) requires the reporting of As, Cd, Cu, Cr, Hg, Ni, Pb, and Zn,3 and the U.S. Clean Air Act Amendments Bill of 1990 lists 11 elements to be of potential concern, namely, As, Cd, Cr, Hg, Ni, Pb, Sb, Be, Mn, Se, and Co.4 Most trace elements in coal are associated with the mineral portion of the coal and considered to have an inorganic association. Others may be intimately associated physically or chemically with the organic matter in the coal, thus having an © 2013 American Chemical Society

organic association. Often, the mode of occurrence is much more important than the actual concentrations because it is the former that dictates the mobilization of these elements during processes, such as pyrolysis, combustion, and gasification, and also the type of elements that could be removed from coal by various separation techniques. The elements whose concentrations increase with an increasing ash content are considered to be associated with the mineral matter and are termed inorganically bound. Al, K, Si, Ti, Sb, As, Be, Cs, Li, Ni, Pb, V, Zn, Rb, Mn, Fe, and most rare earth elements usually follow the ash trend, while some elements, such as Na, Sr, S, Br, and B, are mostly organically bound, and their concentration decreases with the increase of the the ash content. Some elements, such as As and Mn, exhibit a mixed association.5 Pyrite and other iron sulfide minerals have been found to attract trace elements introduced to the coal from anthropogenic and natural sources.6 Studies based on data collated from the U.S.A., Australia, and Great Britain coals show that As, Cd, Co, Cu, Fe, Hg, Mo, Ni, Pb, S, Sb, Se, Ti, W, and Zn occur Special Issue: Impacts of Fuel Quality on Power Production and the Environment Received: March 14, 2013 Revised: June 13, 2013 Published: August 13, 2013 5607

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

reduces slag viscosity and fusion temperature, while the increase of acidic oxides, such as Si and Al oxides, increases viscosity and coal ash melting temperature.18,20,27 Of course, this should be considered carefully because molten ash creates an environment of several compositions, which can react or destroy other component network structures and create unexpected results. Kim et al.18 reported a low fusion temperature for the high quantity of SiO2 and low CaO, where everyone expects a high fusion temperature. Fayalite (Fe2SiO4) formation as result of the high amount of Fe2O3 with a low fusion temperature is reported as the main reason for that. In contrast, Kim et al.20 and Song et al.22,28 reported an increase in the fusion temperature when the amount of inherent or externally added reducing agents (Fe, Ca, Na, and Mg oxides) exceeded a certain amount. The minimizing level (of the fusion temperature) varies from coal to coal. Although chemical coal cleaning can effectively remove most of the minerals and trace elements, their complexity and high incidentals is a huge barrier for commercial development. Dry physical coal beneficiation offers an economic and efficient coal cleaning while avoiding problems associated with wet methods, such as high water consumption and the need for severe coal drying stages prior to energy or liquid fuels production. To the knowledge of the authors, such a fundamental study on the characteristics of the produced clean coal by ADMFB has not been reported elsewhere. In the previous section of our work (10.1021/ef400456n), the effect of operating parameters on the performance of an ADMFB coal separator was discussed in detail and relevant operating conditions that offer the best coal recovery, least ash content, and highest system separation efficiency were introduced. In our current study, which is performed on a Canadian lowash/rank coal, analysis have been performed to study the characteristics of the generated clean coal products and the performance of the system (ADMFB separator) in reducing different kinds of materials, especially hazardous, trace elements from coal, as well as the product fusion and slagging characteristics. For the purpose of this work, several clean coal samples with desirable apparent quality from those experiments (part 1; 10.1021/ef400456n) are accurately characterized. Also, the extent of rejection of selected hazardous volatile trace elements, such as Hg, As, Se, Pb, etc., and nonvolatile trace elements, such as Ag, Ba, Cu, Ni, Sb, Co, Mn, Be, etc. (which are mostly the regulated elements around the world), was investigated. The combustion properties of the clean coal products were also studied using TGA.

in iron sulfides, in particular pyrite form, while Ba, Ca, Fe, and S tend to the sulfates.7 It has been shown that Cu, Mn, Ni, Pb, and Zn can also be associated with silicates/carbonates as well as pyrite.8,9 A positive correlation between Hg and organic sulfur has been reported by some researchers,10,11 and an affinity of Cr and V for clays has been reported by others.7 The effect of demineralization (mostly chemical treatment) and addition of specific inorganic compounds on the reactivity of coal in combustion or pyrolysis has been investigated extensively.12−17 Several criteria have been suggested for characterization of the reactivity. The maximum rate of weight loss (Rmax) and peak temperature (Tmax) are two parameters that have been used by many researchers.12−17 Hanzade et al.12 investigated the effect of chemical demineralization of 25 different lignite coals and observed that reactivity in terms of Tmax increased in 17 samples and decreased in the remaining samples. The increase in the porosity of coal particles because of chemical demineralization and the loss of the catalytic effect of the ash minerals are two competing effects that determine the final outcome of the demineralization on coal reactivity.12,17 Quanrun and colleagues14 used Al2O3, CaO, and K2CO3 as catalysts for the demineralized coal, which resulted in improvement of coal reactivity (Rmax) and lowering activation energy necessary for coal pyrolysis. Katherine et al.15 found CuCl, AgCl, and Cu(NO3)2 as the most effective catalysts, increasing the thermogravimetric analysis (TGA) burnout rate of coal chars among several tested compounds. Therefore, it can be concluded that demineralization or adding metal compounds was found to increase or decrease the coal reactivity depending upon the properties of the studied coals because of the severe heterogeneous nature of coal. Coal ash minerals contain many components with different behaviors when heating to their melting point.18,19 The removal of some ash-forming minerals can affect the slagging and fluxing properties of the beneficiated coal ash and, consequently, the operation of coal conversion units. The slagging and fouling decrease the efficiency of heat-exchange surfaces in conventional coal-firing furnaces, while in slagging gasifiers, where ash is intentionally converted into liquid slag (better operation, control particulate matter emission, and trap trace elements and heavy metals in a unleachable glass phase), to achieve free flux toward the bottom of the gasifier (tapping system), the higher slagging propensity and lower viscosity at the operating temperature are required.18,20,21 For such a purpose, the temperature of the gasifier should be maintained above the fusion temperature of the ash to enable continuous slag tapping.20,22 Several indices and factors, such as ash fusion temperature test (under a reducing environment), base/acid ratio, silica percentage, slagging factor, Fe/Ca ratio, and slagging index, are suggested to predict and evaluate the behavior of solo ash molten slags or the presence of some additives.18,19,23−26 Severe slagging is expected when the reducing ash fusion temperature is below 1350 °C.23,24 The lower difference of flow and initial deformation temperature (obtained from the ash fusion test), result in thinner and adhesive ash deposit on the reactor surface.18,20 Generally, the fusion temperature in a reducing atmosphere is equal or less than that in the oxidation atmosphere, where the decrease in the Fe content of the ash reduces the difference.20 Slag viscosity (usually) less than 250 P is suggested to avoid problems of slag tapping from the gasifiers.18,20,26 It is well-demonstrated in the literature that the increase of some basic oxides, such as Fe, Ca, and Na oxides,

2. MATERIALS AND METHODS A series of experiments was designed and conducted in a batch ADMFB unit with 40 cm height and 20 cm inner diameter to study the effect of the main operating parameters, namely, air superficial velocity, bed height, separation time, and particle size, on the performance of the ADMFB separator and find the optimum conditions that provide maximum ash removal or recovery of the organic material to the clean coal product. The detail of the setup and experiments can be found in the first part of this work (10.1021/ef400456n). A low-ash lignite coal (Boundary Dam lignite coal) was used in beneficiation tests. Most of the tests were performed with the particle size range of 2.8−5.6 mm, but some tests were also performed with two different particle size ranges to investigate the effect of the feed particle size. The properties of the feed samples used for beneficiation tests are presented in Table 1. 5608

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

column. Particles were collected from 5 cm layers starting from the top of the bed to the bottom. The very top layer is referred to as L1 (low-density region), where the number in the suffix is increased by moving from top layers to the bottom; e.g., the third layer (10−15 cm) is called L3. As the panels show, by moving from the top layer, L1, to the bottom layer, L4 (highdensity region), C, H, N, O, and HHV of the particles decrease, while S and ash increase. The change is more pronounced between L3 and the bottom of the bed (L4). It means that a great deal of ash and ash-related component accumulation occurs right at the bottom. The ash reduction profile along the bed depth determines the best location for the side streams of separation product removal that gives the desirable quality of the products. As discussed, both organic and mineral types of S can occur in coal. Therefore, the increase in the S content of the L4 (33% increase with respect to the feed sample) could be attributed to the occurrence of S carrying minerals in this zone, because these minerals are heavy and can easily sink into the fluidizedbed depth, to the lower zone. 3.2. Mercury Content. The mercury content of the feed samples was measured to be 92, 94, and 67 ppb (mg/ton) for fine, middle, and coarse size fractions, respectively. The mercury rejections for different samples relative to their feed samples are provided in Table 4. An increase of 10% is detectable for fine clean coal products. An observed strong association of the mercury content with the mineral matters of coal could be the result of fine minerals occurring in L1. These very fine mineral particles (because of comminution) are easy to be carried by the fluidizing air in the bed because the operating air velocity is higher than their minimum fluidization velocity. For the middle and coarse size fractions, a significant rejection of the mercury has been achieved. The minimum detected mercury content for this size fraction is 48 ppb, equal to 48.6% rejection in the mercury content compared to its relevant feed. The mercury rejection obtained for the coarse particles was in the range of the middle size fraction. Studies have shown that mercury in the coal is mostly associated with the mineral impurities (mainly with pyritic sulfur).32,33 Although pyrite is not a significant component of the Canadian coals, a direct relation between the rejection of run of mine (ROM) mineral matters and hazardous elements as a result of cleaning processes has been revealed.33−35 The independency of the Hg content to the coal particle size is also reported in some studies.35 The ash and corresponding Hg contents of coal samples were used to establish a correlation, as shown in Figure 2. As this figure shows, a good linear correlation (a positive trend) between ash and Hg content with a regression coefficient (R2) of 0.9 is obtained. Considering the mercury

Table 1. General Properties of Feed Coals Used in the ADMFB Separator sample name

particle size (mm)

average ash content (%)

fine feed middle feed coarse feed

1−2.8 2.8−5.6 +5.6

14.4 12.5 11.7

A number of beneficiated samples from ADMFB separator tests were selected for the detailed characterization. These samples include two clean coal products, which exhibited minimum ash content (B and C), two tests that offered maximum organic material recovery (D and E) from the middle particle size fraction, and two tests with fine and coarse particle sizes (A and F, respectively). Table 2 summarizes the general characteristics of these samples. All selected clean coal products were ground to finer than 0.2 mm and analyzed with TGA (Q600 TGA-DSC, TA Instruments, New Castle, DE), CHNS analyzer (Vario Micro), direct mercury analyzer (DMA-80), inductively coupled plasma−mass spectrometry (ICP− MS, Perkin-Elmer’s Elan 6000 ICP−MS), EDAX energy-dispersive Xray fluorescence (XRF) microprobe system (rhodium X-ray source), Barnstead Thermolyne furnace (muffle furnace model 6000), and Preiser automated ash fusion furnace (this test was performed by Birtley Coal and Minerals Testing Division, Calgary, Alberta, Canada) to study various properties of these products and their ashes. The analyses were repeated 2 or 3 times to ascertain the reproducibility of the analysis results.

3. RESULTS AND DISCUSSION 3.1. Ultimate (CHNS) Analysis. The ultimate analysis and higher heating value (HHV) of six clean coals as well as their feed samples are presented in Table 3. Each measurement was repeated at least 3 times to ascertain the accuracy and consistency of the measurements. The oxygen (O) content of the samples was calculated on the basis of CHNS data. The HHV was calculated using proper literature correlations.29 A review of Table 3 reveals that, by decreasing the mineral content of the clean coals, the proportion of C, H, N, and S components of the clean products slightly increases, regardless of the particle size of the samples. This is an indication of the organic origin of those components.30,31 Among the four products of the middle size range, the tests with the least ash content (B and C) show a greater increase of N, C, S, and HHV (on average 8.3, 3.7, 4.9, and 3.7% increase for N, C, S, and HHV, respectively) compared to the high recovery tests (D and E). This is not surprising because the high recovery products are more similar to the feed sample because a lower amount of materials are rejected from the top (L1 layer) to the lower layers throughout the separation process. Panels a and b of Figure 1 represent the variation of different components of CHNS analysis and HHV for a middle size fraction sample gathered along the depth of the ADMFB

Table 2. General Specifications of the Selected Beneficiated Samples for Characterization characterization name

beneficiation test (number)

particle type

ash content (%)

ash removal with respect to feed (%)

organic material recovery (%)

A B C D E F

1a 9b 10b 11b 16b 5a

fine middle middle middle middle coarse

13.25 9.31 10.10 10.39 9.86 8.77

9.4 17.2 18.9 14.3 20.2 22.1

68.36 72.66 70.73 97.36 89.83 87.72

a

More information available in Table 3 of part 1 (10.1021/ef400456n). bMore information available in Table 2 of part 1 (10.1021/ef400456n). 5609

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

Table 3. Ultimate Analysis and HHV of the Coal Samples sample name

coal ash (%)

N (%)

C (%)

H (%)

S (%)

O (%)

HHV (MJ/kg)

fine feed A middle feed B C D E coarse feed F

14.63 13.25 12.58 9.31 10.10 10.39 9.86 11.25 8.77

0.935 0.950 0.984 1.067 1.063 1.033 1.027 0.925 0.950

55.19 55.98 56.62 58.69 58.74 57.24 57.72 56.85 58.56

4.08 4.13 4.19 4.24 4.22 4.30 4.30 4.31 4.42

0.608 0.589 0.604 0.651 0.617 0.591 0.609 0.599 0.612

24.56 25.10 25.32 26.05 25.26 26.44 26.48 26.07 26.69

21.27 21.58 21.87 22.64 22.71 22.14 22.32 22.03 22.75

Figure 1. Component and parameter distribution along the bed depth: (a) N, H, and S and (b) C, O, ash, and HHV.

coal particle sizes shows the same trend as for the middle size in terms of hazardous element rejection, except for Ni. In terms of the particle size effect, for most of the elements, cleaning the fine particles was more effective in reducing the hazardous elements than middle or coarse particles (panels a and b of Figure 3). This could be attributed to the higher degree of liberation of the ash-forming minerals from the organic phase when dealing with fine particles. Surprisingly, high accumulation of Ni was observed for the coarse clean coal product, where the occurrence of a highly Ni-enriched particle in the L1 layer of the bed or lack of the proper particle liberation can cause that. A decrease in hazardous elements of the clean coal products is an indication of a strong positive affinity of these elements with ash-forming minerals in the coal. This trend has also been reported by Finkelman36 for coals with ash contents of greater than about 5%. Generally, Cu, K, and Pb exhibit the highest degree of removal, which is an indication of their strong bonding with coal mineral matter. Low-rank coals, such as the lignite used here, usually contain more organically bound elements because of being enriched with oxygen-bearing functional groups, such as carboxylic acid (−COOH) and phenolic hydroxyl (−OH) groups, which are lost with increasing the coal rank.37,38 The −COOH group readily participates in ion-exchange reactions and the formation of organometallic complexes, commonly reported in lignites.5,38,39 The different behaviors of As, Se, Sb, and Co with a decreasing ash content of the coal (panels a and b of Figure 3) reveal the presence of some organically bound fraction of these elements or, in another term, the strong association of these elements with the organic phase. 3.4. Ash Composition. The composition of the feed samples and changes imposed on the beneficiated product ashes corresponding to their relevant feeds are presented in Table 5. The ashes are produced by Barnstead Thermolyne muffle furnace (model 6000) according to the ASTM D3174

Table 4. Mercury Content Analysis Results sample name

Hg content (ppb)

Hg rejection (%)

A B C D E F

101.2 51.8 62.5 49.3 58.5 41.6

−9.9 (increase) 45.1 33.7 47.7 48.6 37.4

Figure 2. Association of Hg and the ash content of the Boundary Dam coal.

rejection results in Table 4 and Figure 2, it can be concluded that removing the coal ash with ADMFB can efficiently reduce the Hg content of the coal. 3.3. Trace Elements. The ICP−MS analysis was performed on the clean coal products to study the change in the trace element distribution because of the beneficiation. The results are provided in panels a and b of Figure 3. As seen from both panels, the amount of most of the selected hazardous elements decreases after treating with ADMFB separator. Negative values in graphs are indicative of the rejection in the trace element contents. The effect of beneficiation on both fine and coarse 5610

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

Figure 3. Trace element removability graphs because of beneficiation of coal by ADMFB: (a) changes imposed on the middle size product and (b) changes imposed on fine and coarse products.

Table 5. Changes in Chemical Composition of the Ashes

a

sample name

Na2O (%)

MgO (%)

Al2O3 (%)

SiO2 (%)

SO3 (%)

CaO (%)

Fe2O3 (%)

P2O5 (%)

fine feed changes in Aa middle feed changes in B changes in C changes in D changes in E coarse feed changes in F

7.93 9.25 9.06 22.66 17.65 10.51 14.22 9.69 22.42

3.48 −7.22 4.06 2.51 −1.67 −1.92 −7.20 3.76 19.24

19.27 −2.53 17.65 −3.03 −3.44 −1.09 −0.54 17.41 −14.23

31.74 3.85 26.18 −11.93 −11.53 −7.25 −2.29 25.14 −31.67

15.79 −6.64 17.92 −5.13 0.19 1.34 1.66 19.29 20.02

15.01 −1.02 17.14 11.62 10.85 5.19 −0.87 16.47 20.42

3.02 6.29 3.77 6.83 3.98 2.60 −5.27 3.78 18.47

1.64 −4.97 2.25 9.32 4.81 0.09 −7.74 2.30 −4.47

Negative signs indicate a decrease with respect to the feed sample.

and Mg could be associated with the mineral matter of the coal samples rather than the possibility of being organically bound between them and the organic phase. The SO3 and P2O5 contents do not appear to present any trend regarding the feed particle size. On the other hand, for the middle and fine particle sizes, S shows no correlation with Fe, indicating that more S is occurring as organic type than mineral type41 in the studied coal. Although the organic origin of S is supported by the CHNS results, this conclusion is not made with higher precision here because there is no clear negative correlation of S and ash reduction of products. As discussed before, beneficiation can change the composition of the clean coal ash by removal of various mineral components in the ash, partially or completely, depending upon how bound they are with the coal phase. The slagging and fouling tendency of some feed and beneficiated samples are compared in Table 6 in terms of some commonly used indices, such as base/acid ratio (B/A), slagging factor, silica percent, and Fe/Ca ratio, as described in eqs 1−4, respectively.

(Standard Test Method for Ash in the Analysis Sample of Coal and Coke from Coal) method. Just oxides with high assay (grade) are presented, and low-assay compounds are neglected. Ash analysis results are more reliable than the analysis of whole coal because, for most of the compounds (non-volatile), higher enrichment is obtained when the dominant organic part is burnt out.40 When the ICP−MS and XRF results are compared, this point should be considered that XRF directly measures the total amount present, while ICP−MS only detects the elements that are totally or partially leachable into the acid.41 Therefore, the results for the elements that can be organically or covalently bound to the coal phase (e.g., organic S or Al, Cl, Fe, trace elements, etc.) should be interpreted with care.38,41 The characterization results in Table 5 indicate that the Na2O content of all products, regardless of the feed particle size, has increased at least 9%. The CaO and Fe2O3 contents of the products have usually increased to different extents. Considering the fact that all six products have lower amounts of ash minerals than their corresponding feed samples and also the increasing trend (negative affinity) of the three mentioned elements, it can be concluded that Na, Ca, and Fe are generally associated (bound) with organic materials than the ash-forming minerals.38,41 Such inorganic elements can be bound to the organic phase directly or ion exchange, particularly for low-rank coals, as salts of carboxylic acids (−COOH groups), as discussed in the previous section.37−39,41 The Al2O3, MgO, and SiO2 contents of the ashes show different levels of reduction when compared to feed samples. The three mentioned elements are the basic components of mainly clay minerals (illite, smectite, and kaolinite), which are expected to be abundantly distributed in the coal seams (either syngenetic or epigenetic) or are introduced from hanging wall or floor rocks during the mining operations. Generally, Si, Al,

Table 6. Calculated Slagging Indices

5611

sample name

B/A ratio

slagging factor (%)

silica percent

Fe/Ca ratio

fine feed A middle feed B C D E coarse feed F

0.57 0.58 0.61 0.94 0.92 0.85 0.79 0.78 1.25

6.62 6.19 7.23 11.37 11.80 11.07 10.44 10.94 20.47

59.61 60.76 58.63 45.76 46.25 48.41 51.25 51.15 37.36

0.20 0.22 0.21 0.21 0.21 0.21 0.20 0.23 0.23

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

Table 7. AFT Test Resultsa reducing atmosphere

a

oxidizing atmosphere

sample name

IDTb

STc

HTd

FTe

IDT

ST

HT

FT

slagging index

fine feed A middle feed B C D E coarse feed F

1102 1107 1091 1131 1144 1091 1136 1091 1081

1148 1147 1141 1161 1167 1147 1160 1162 1158

1157 1152 1155 1166 1174 1158 1163 1165 1164

1244 1195 1213 1196 1200 1164 1171 1202 1184

1149 1160 1150 1197 1191 1186 1169 1165 1223

1348 1357 1281 1413 1422 1411 1405 1388 1394

1362 1369 1331 1418 1424 1414 1409 1402 1402

1423 1423 1400 1431 1432 1422 1425 1433 1424

1113.0 1116.0 1103.8 1138.0 1150.0 1104.4 1141.4 1105.8 1097.6

All numbers are in °C. bInitial deformation temperature. cSoftening (spherical) temperature. dHemispherical temperature. eFluid temperature.

Fe O + Na 2O + K 2O + MgO + CaO B = 2 3 A SiO2 + Al 2O3 + TiO2 slagging factor = silica percent =

Fe O Fe = 2 3 Ca CaO

B S(%) A SiO2 × 100 Fe2O3 + SiO2 + MgO + CaO

3.5. Ash Fusibility. Ash fusion temperature (AFT) test results are presented in both under reducing and oxidizing environments. An oxidizing environment is generated when coal is burnt with an abundant amount of oxygen at lower temperatures (around 800 °C), such as in conventional burners, while the reducing atmosphere occurs in gasifiers where the operating temperature is high and oxygen is not abundantly available. As seen, AFT in a reducing environment has decreased for all beneficiated samples. Generally, a higher slagging propensity is expected when AFT in a reducing environment is less than 1250 °C. This reduction in AFT for beneficiated products is favorable for slagging and fluxing needs in slagging gasifiers. Lower AFT for beneficiated products results in faster slagging as well as lower molten slag viscosity at a certain operating temperature in the gasifier. A decrease in AFT for B and C (low-ash products) is less than that for D and E (high-recovery products). AFT under an oxidizing environment is higher than its equivalent AFT under a reducing environment. According to Table 7, oxidizing AFT for middle products has increased in comparison to their corresponding feed samples while remaining constant or decreasing for fine and coarse products, respectively. An increase in oxidizing AFT for the beneficiated products leads to reduced slagging. Slagging indices (eq 5) calculated on the basis of the reducing AFT for each ash are also presented in Table 7. As seen, the slagging factor for all feeds and products is less than 1170 °C. For such slagging factors, severe or high slagging as well as low flux viscosities are expected.23,24

(1)

(2)

(3)

(4)

Ash XRF results are used to calculate these indices. It can be seen that the B/A ratio has increased for all beneficiation products. Usually, the B/A ratios lower than 0.7 result in higher slagging propensity, while in comparison to that, for the ratios higher than 0.7, lower slagging tendencies are expected.23,24 The increase of B/A is more for middle size products (in comparison to the corresponding feed) than the fine size fraction, where the bigger change in the B/A ratio occurred for course samples. Also, the increase of the B/A ratio for B and C is more than that for D and E. The slagging factor has also increased for all beneficiated samples. The slagging factors of the beneficiated samples follow the same trend as the B/A ratio, and a bigger change occurred for the coarse particle size. Ashes with slagging factors higher than 2.6 are expected to exhibited good (increased) slagging behavior.24 The silica percent is a good indication of the slagging and fusion properties of the coal ashes in burners and gasifiers. A lower silica percent (less than 65%) is usually considered as a lower fusion temperature and viscosity for the molten slag from the ashes. This has been presented in Table 6. Except for the fine product, this index decreased for all five beneficiated samples. The coarse product experienced the highest decrease. All analyzed feed samples have a silica percent less than 65%, meaning a high tendency to produce earlier melting lowviscosity slags, where products that decreased silica percent even more intensifies that. For low-ash middle size products (B and C), the silica percent decreased more than the highrecovery products. All Fe/Ca ratios for feed or beneficiated samples (Table 6) fall in a narrow range of 0.2−0.23, regardless of the particle size. Medium to high slagging tendency is expected when this ratio falls between 0.3 and 3. The calculated indices presented in Table 6 indicate that, in general, the physical beneficiation of coal samples increases the slagging propensity of the clean coal ashes, which is an advantage when these products are fed into the slagging gasifiers.

slagging index =

4IDT + HT 5

(5)

Viscosity of all selected products and their relevant feed samples was calculated using an Urbain viscosity model at 1250 °C under both oxidizing and reducing environments. Urbain correlation is one of the most widely used slag viscosity models.42−46 This model is largely dependent upon the SiO2 content of the slag. The predicted viscosities are presented in Table 8. As seen, the reducing environment viscosity is less than the oxidizing environment viscosity for any individual sample. This difference is not much, because the Fe content of the ashes is low, resulting in closer FeO or Fe2O3 mole fractions when considering different environments. Considering Table 8, viscosity of the fine product for both oxidizing and reducing environments has increased. This can be attributed to the decrease of CaO or increase of SiO2 in the fine product compared to its feed. For both middle and coarse 5612

dx.doi.org/10.1021/ef400457y | Energy Fuels 2013, 27, 5607−5616

Energy & Fuels

Article

Table 8. Molten Ash Slag Viscosities at 1250 °Ca

a

sample name

reducing environment

oxidizing environment

fine feed A middle feed B C D E coarse feed F

76.49 79.56 66.56 30.98 31.8 36.28 40.4 40.6 19.12

82.36 86.32 71.53 33 33.85 38.75 43.12 43.5 20.31

in iron oxide and a decrease in Si and Al in ash analysis, which results in a decrease in viscosity and an increase of slagging tendencies. 3.6. Reactivity. The TGA study was carried out to investigate the reactivity of the beneficiated coal samples toward combustion. The maximum rate of weight loss (Rmax), determined by differential thermogravimetry (DTG) (first derivative of the TGA curve), and the peak temperature (Tmax), defined as the temperature corresponding to the Rmax (%/min), are taken as measures of the coal reactivity, where higher Rmax and lower Tmax show high coal reactivity.12−17,47,48 For the non-isothermal TGA experiments,