Mineralogical and Chemical I Composition of Inorganic Matter from Marketed Illinois Coals llham Demir, Randall E. Hughes, Philip J. DeMaris, and Rodney R. Ruch Illinois Minerals 1 21 2001 George H. Ryan, Governor Department of Natural Resources Brent Manning, Director ILLINOIS STATE GEOLOGICAL SURVEY William W. Shilts, Chief Natural Resources Building 615 East Peabody Drive Champaign, IL 61 820-6964 Mineralogical and Chemical Composition of Inorganic Matter from Marketed Illinois Coals 1 1 ham Demir, Randall E. Hughes, Philip J. DeMaris, and Rodney R. Ruch Illinois Minerals 121 2001 George H. Ryan, Governor Department ot Natural Resources iQtth^ Brent Manning, Director VV& ILLINOIS STATE GEOLOGICAL SURVEY \l$\ I J William W. Shilts, Chief CttlN^ Natural Resources Building «» G^ 1 615EastPeabodyDrive " Champaign, IL 61 820-6964 (217)333-4747 ACKNOWLEDGMENTS We thank R. A. Cahill, Y. Zhang, and J. D. Steele for chemical analyses of the samples. We also thank W. R. Roy and R. A Cahill for critically reviewing the manuscript. This study was supported, in part, by grants from the Illinois Department of Commerce and Community Affairs through the Illinois Coal Development Board and the Illinois Clean Coal Institute. Editorial Board Jonathan H. Goodwin, Chair Michael L. Bamhardt Anne L. Erdmann B. Brandon Curry David R. Larson Heinz H. Damberger Donald G. Mikulic William R. Roy Printed by authority of the State of Illinois/2001/600 © Printed with soybean ink on recycled paper CONTENTS ACKNOWLEDGMENTS ABSTRACT INTRODUCTION RELATED RESEARCH Inorganic Matter in Coal Coal Combustion Residues Chemical and Mineralogical Analyses of Coal and CCRs OBJECTIVES EXPERIMENTAL PROCEDURES Samples and Sample Preparation Chemical Analysis of Coal and CCR Samples Low-temperature Ashing of Coal Samples Water Leaching of LTA Samples Mineralogical Analysis of LTA Samples Mineralogical Analysis of CCRs and Limestone RESULTS AND DISCUSSION Mineralogical Composition of Marketed Coals Relationships between Minerals and Minor and Trace Elements Combustion Behavior of Minerals in Coal Emission of HAPs from Power Plants Partitioning of HAPs between Fly Ash and Bottom Ash Temporal Variations in the Characteristics of Marketed Coals SUMMARY AND CONCLUSIONS REFERENCES 2 2 3 3 5 5 5 6 7 7 7 8 8 8 10 11 15 22 23 25 26 FIGURES 8 9 10 11 12 US production and use of CCRs for the year 1 998 3 Coal regions of Illinois 5 Average mineral content of 35 marketed coals 10 Major mineral composition of mineral matter from 35 marketed Illinois coals 10 Clay mineral composition of the mineral matter from 35 marketed Illinois coals 1 1 Relationships of r 2 > 0.5 between elements and mixed-layered illite/smectite in marketed Illinois coals 1 1 Relationships of r 2 > 0.5 between elements and mixed-layered illite/smectite plus illite in marketed Illinois coals 16 Relationships of r 2 > 0.5 between elements and kaolinite in marketed Illinois coals 17 Relationship between quartz and Mg in marketed Illinois coals 17 Calcite-Mn and calcite-Zn relationships in marketed Illinois coals 18 Relationship between pyrite and Hg in marketed Illinois coals 18 Mineralogical composition of CCRs from three types of plants (FBC, PC, CYC) in Illinois 19 TABLES 1 Amounts and descriptions of samples of coal and CCRs from three types of power plants 6 2 Mineralogical composition of marketed Illinois coals and CCRs 9 3 Chemical composition of 35 marketed Illinois coals from preparation plants and of feed coals and combustion residues from four selected power plants 12 4 Mineral-element correlations in marketed Illinois coals 15 5 Concentrations of Al and 1 5 elements of environmental concern in coal, CCRs, and limestone samples from the four power plants and related combustion mobility and relative enrichment values 20 6 Mobilities of HAPs during combustion 21 7 Relative distribution of HAPs among combustion ashes 23 8 Temporal variations in the mineralogical and chemical composition of marketed coals from four selected Illinois preparation plants 24 ABSTRACT The mineralogical and chemical compositions of marketed Illinois coals and coal combustion residues (CCRs) were investigated to obtain a better understanding of the relationship of these materials to air pollution and ash formation at coal-fired power plants. The coal samples were collected from 35 coal preparation plants in Illinois. The samples of CCRs (fly and bottom ashes), along with samples of feed coals, were collected from a fluidized bed combustion (FBC) plant, two cyclone combustion (CYC) plants, and a pulverized coal combustion (PC) plant. Average (mean ± standard deviation) abundances of minerals in the marketed coals were kaolinite, 2.07 ± 0.56%; quartz, 2.06 ± 0.78%; mixed-layered illite/smectite, 1 .57 ± 0.70%; pyrite, 1 .56 ± 0.50%; illite, 1 .29 ± 0.46%; calcite, 0.84 ± 0.78%; chlorite, 0.18 ± 0.13%; marcasite, 0.13 ± 0.12%; plagioclase, 0.07± 0.05%; and K-feldspar, 0.01 ± 0.01 %. Nineteen minor and trace elements positively corre- lated with some of the minerals. The correlations that involved elements of environmental concern were Mn-calcite, F-illite and F-illite/smectite, Th-tllite and Th-illite/smectite, Th-kaolinite, Cr-illite/ smectite, and Hg-pyrite. A comparison of the mineralogical composition of coals with CCRs sug- gested that combustion at the power plants partially converted quartz to silica glass. The clay minerals and feldspars were converted largely to aluminosilicate glass and, to a much lesser extent, to mullite. Pyrite and marcasite were converted to hematite, magnetite, and amorphous Fe oxides. Calcite was converted to anhydrite, gypsum, lime, and portlandite; some of the lime and portlandite was converted back to calcite through reaction with atmospheric C0 2 .The minerals derived from calcite formed 30% and 71% of the FBC fly ash and bottom ash, respectively. The CCRs from the CYC and PC plants contained mostly amorphous material (78-98%). Other minerals in the CCRs from the four power plants were magnetite (0-4.95%), quartz (0.008-8.44%), mullite (0-5.09%), hematite (0-3.03%), and calcite (0-1 .95%). The combustion behaviors of 15 elements (As, Be, Cd, Co, Cr, F, Hg, Mn, Ni, R Pb, Sb, Se, Th, and U) of environmental concern in coal were estimated by comparing the analytical data from the feed coals and CCRs from the power plants. These elements were partly retained in the CCRs and partly mobilized and airborne upon combustion. Fluorine and Mn mobilities were greatest (>50% of the original amounts in the feed lost) at the FBC plant, and F, Hg, and Se mobilities were greatest at the CYC and PC plants. The majority of the other elements were less mobile (<25%). The mobilized portions of the elements may not have been entirely emitted into the atmosphere; they may have been partially condensed on hardware at the cool side of the power plants. The concen- trations of the 15 trace elements were greater in the fly ashes than in the bottom ashes, with a few exceptions. Some of the elements could be commercially produced from the CCRs. Samples collected at four selected coal preparation plants at a 4-year interval showed some variations in the composition of the coals over time. These results give a general idea about the limited reliability of using outdated coal-quality data from a given mine for assessing the environ- mental and economic impacts of burning the coal currently produced from the mine. INTRODUCTION The industrial use of coal is greatly affected by the amount and variety of its inorganic content. The inorganic matter is the main source of pollutants emitted to the atmosphere during coal combustion or leached from wastes from various coal utilization processes. The inorganic matter in the coal, however, may also be a source of useful by-products. Data on the mineralogical composition of coal and CCRs have broad applications in studies relating to commercial uses of Illinois coals and of the residues from cleaning, combustion, and conversion of these coals. Interpreting combined mineralogical and chemical data helps identify the modes of occurrence of toxic elements in coal; such information is essential for designing coal cleaning or post-combustion control methods to meet the requirements of current or future environmental regula- tions. Improved understanding of the fate of minerals in different types of boilers also increases the ability to design improved methods to predict slagging, fouling, and ash properties from the mineral- ogical and chemical composition of feed coal. Ash fusion temperatures, slagging, and fouling are strongly dependent on the mineralogical composition of coal and, therefore, cannot be accurately estimated from chemical analysis of coal or ash alone. The minerals in coals melt over a range of temperatures, which reflects the refractoriness of each type of mineral. The difference between the ash fusion temperature predicted from bulk chemistry and the actual or correct value increases as the partitioning of elements among different minerals increases. The CCR data can also help generate new ideas about increasing their commercial utilization. RELATED RESEARCH Inorganic Matter in Coal The inorganic matter content of run-of-mine coal is highly variable and depends on geological conditions and the mining techniques used. Inorganic matter in coal is present mostly as discrete minerals with five different modes of occurrence: (1 ) microscopically disseminated inclusions within macerals; (2) layers or partings of variable thickness; (3) nodules, including lenticular and spherical concretions; (4) fissures, including cleat and other fracture and void fillings; and (5) megascopic rock fragments found within the coal bed because of faulting, slumping, or related disturbances and from intentional or accidental mining of the roof or floor strata (Harvey et al. 1 983, Harvey and Ruch 1 986). Physical coal-cleaning processes normally remove large, discrete grains and layers of inorganic matter more efficiently than they remove finely disseminated minerals. Mineralogical compositions of inorganic matter in channel and core samples of Illinois coals were investigated by Rao and Gluskoter (1 973), Ward (1 977), and Harvey et al. (1 983). Minerals that commonly occurred in measurable quantities in these Illinois coal samples included illite, mixed-layered illite/smectite, kaolinite, pyrite, quartz, and calcite. Minerals with rare occurrences and trace abundances in the samples included chlorite, feldspars, zircon, sphalerite, dolomite (ankerite), siderite, barite, anhy- drite, gypsum, rosenite, melanterite, coquimbite, jarosite, hematite, rutile, and apatite. Inorganic minerals host the elements of environmental concern in coal. However, there are a few exceptions; for example, S can be part of the organic structure, and CI can partly be adsorbed on the organic matter in exchangeable ionic form. Most elements of environmental concern in coal are categorized as trace elements (generally <200 mg/kg). Gluskoter et al. (1 977) and Harvey et al. (1 983) reported on the trace element contents of a large number of mostly channel and core samples of Illinois coals. Demir et al. (1 998) used principal component analysis to determine the modes of occurrence of environmentally critical elements in channel samples of Illinois coals. With respect to coal utilization, the National Committee for Geochemistry (National Research Council 1 980) identified three categories of elements of environmental concern: (1 ) those of great- est concern — As, B, Cd, Pb, Hg, Mo, and Se; (2) those of moderate concern — V, Cr, Ni, Cu, Zn, and F; and (3) those of minor concern — Li, Na, Sr, Ba, Mn, Co, Ge, CI, Br, Ra, Po, Rn, Th, and U. These categories were based upon known toxicity, levels of occurrence of each element in coal, and anticipated mobility upon combustion or upon disposal of ash. Sixteen elements in coal (As, Be, Cd, CI, Co, Cr, F, Hg, Mn, Ni, R Pb, Sb, Se, Th, U) are among the 1 89 hazardous air pollutants (HAPs) mentioned in the 1 990 Clean Air Act amendments (U.S. Public Law 1 01 -549, 1 990). It should be pointed out that some of the elements in the National Research Council list were not considered HAPs by the amendments. Partitioning of HAPs and other elements among the CCRs and flue gas is highly variable (Swaine 1 990, Clarke and Sloss 1 992, Wesnor 1 993, Davidson and Clarke 1 996) because of the variation in the types and operational conditions of combustion units and in the chemical properties and modes of occurrences of the elements. Analytical errors and difficulties in obtaining representative samples also add to the variability of data on the retention or emission of potentially toxic elements from power plants. The HAP provisions of the 1 990 amend- ments to the Clean Air Act focus on non-utility industries, including metals, petrochemicals, and papermaking (Moore 1 994). A risk analysis by the U.S. Environmental Protection Agency (USEPA 1998) concluded that, for now, only Hg emission from the coal-fired electrical utilities is of potential health risk and thus requires further investigation; a final regulation on Hg will be issued by 2004. fly ash bottom ash boiler slag FGD material Figure 1 US production and use of CCRs for the year 1998 (plotted using data from ACAA 1999). FGD, flue gas desulfurization. Coal Combustion Residues Utilities in the U.S. annually generatel 08 million short tons of CCRs, 58% of which is fly ash (fig. 1 ). Current and potential commercial uses of CCRs include treatment of soil and of mine and industrial wastes; admixture in cement, concrete, and grout; making bricks and other ceramic products; mineral filler; fill materials in civil engineering projects including road base/subbase; blasting grit; roofing granules; making wallboards; snow/ice control; and extraction of valuable materials (Clarke 1 992, DeBarr et al. 1 996, Dreher et al. 1 996, Hughes et al. 1 996, Wirtz and Bukowski 1 996, Roy et al. 1 997, ACAA 1 999, Barbieri et al. 1 999, Dhir and Jones 1 999, Niewiadomski et al. 1 999, Sloan et al. 1 999). However, only about 29% of the total CCR generated in the U.S. is used for commercial purposes (fig. 1 ); the rest is disposed of on land or in abandoned mines. Chemical and Mineralogical Analyses of Coal and CCRs Methods for determining the chemical composition of coal and CCRs have been well established at the Illinois State Geological Survey (Gluskoter et al. 1 977, Harvey et al. 1 983, Demir et al. 1 994). Methods for mineralogical analysis of these same materials, however, vary greatly in their useful- ness for solving various problems associated with coal use. Application of an improved x-ray diffrac- tion (XRD) method for mineralogical analysis of coal and CCR samples was an important task of this study. Therefore, the mineralogical analysis methods are discussed in considerable detail. The methods of analyzing minerals in coal are based on XRD analysis, chemical analysis, or a combination of the two. Methods based solely on conventional XRD analysis contain large errors, as do those based solely on chemical analysis (called normative methods). Errors in determining the mineralogy of a coal based solely on the chemical analysis generally result from assigning the chemical elements to the wrong normative minerals. Another shortcoming of normative methods is that they are costly and generally lack procedures for estimating errors (Slaughter 1 989). As a result, most investigators have favored methods that combine XRD, chemical analysis, and sometimes other procedures such as scanning electron microscopy. An interlaboratory study of quantitative coal mineral analysis by computer-controlled scanning electron microscopy indicated that the mineralogical composition of coal cannot be determined precisely using this technique alone (Galbreath et al. 1 996). Beginning in the 1 970s, investigators at the Illinois State Geological Survey began to apply combined XRD and chemical methods for analysis of the non-clay mineral fraction of coals. It is important to note that the methods described in the literature generally lack provisions for the analysis of some important minerals, particularly marcasite, mixed-layered illite/ smectite, and mixed-layered kaolinite/expandables. There is extensive literature relating to the determination by XRD of mineral matter in coal. The following developments were especially relevant to this present study: 1 . Development of a procedure for removing the combustible material from coal at low tempera- tures, called low-temperature ashing (Gluskoter 1 965). 2. Development of an accurate and precise method for the determination of pyritic sulfur in coals (ASTM1995). 3. Development of methods for the quantitative determination of minerals by XRD (Snyder and Bish1989). 4. Development of NEWMOD® (Reynolds 1 985) and its application to the quantitative analysis of clay minerals (Hughes and Warren 1 989, Reynolds 1 989, Hughes et al. 1 994, Moore and Reynolds 1997). The mineral concentrations in bulk coal samples often are too small to be accurately and precisely determined. Therefore, XRD of the low-temperature ashes (LTAs) of coals has been the principal method for mineralogical determinations for more than 30 years. The method yields a virtually certain identification of the minerals in a sample. In addition, variations in the relative intensity of XRD peaks for some minerals commonly can be used to estimate their chemical composition. The precision of different versions of XRD analysis varies widely, as do claims of their accuracy (Snyder and Bish 1 989, Moore and Reynolds 1 997). However, as a result of refinements during the 1980s, precise and accurate determination of non-clay minerals such as pyrite (and marcasite), quartz, dolomite, and calcite became possible (Harvey et al. 1 983, Demir and Harvey 1 990). A later study established analytical procedures that meet the requirements of most investigators for mineralogical analysis of inorganic matter in coal (Kruse et al. 1 994). These procedures reduced or eliminated the following errors associated with historical XRD procedures: 1 . Preferred orientation error. Two advances have been made that largely eliminate the error resulting from preferred orientation of mineral particles. The first is to grind the sample in a manner that creates a more random orientation of the crystallites. The McCrone micronizers® in our laboratory yield powders with crystallite sizes in the optimal 5-um to 1 0-um range. The second advance involves using side-loading powder holders and the capability to spin the samples during exposure to the x-ray beam to minimize preferred orientation effects. The typical practice of the past decade has been to use a random powder pack for non-clay mineral analysis and a bulk smear sample for analysis of clay minerals (Hughes et al. 1 994). Historically, XRD analysis of clay minerals has been run on the segregated <2-um fraction of a sample. This practice was developed partly because instruments at the time gave low peak intensities for most clay minerals and partly because clay minerals were thought to be essentially associated with the finest particle fractions, <2 urn by definition. In reality, the clay minerals in coal occur in a wide range of particle sizes. For example, illite/smectite crystallites commonly are <1 urn, while some kaolinite crystallites exceed 50 urn. As a result, errors in excess of an order of magnitude can occur from XRD analysis of samples prepared as <2-um aggregates. This error has been eliminated by grinding the clay minerals to approximately the same size and by analyzing a bulk sample that includes all of the particles (Hughes and Warren 1 989, Hughes et al. 1 994). 2. Standard mineral error. The standard mineral error results from the measurement or calculation of a reference intensity for a "standard" mineral (Russell and Rimmer 1 979, Hughes et al. 1 994, Moore and Reynolds 1 997). In the past, mineralogists analyzed standard samples with the highest purity possible and refined their procedure whenever a better standard was located. For clay minerals, Reynolds (1 985) developed a computer program (NEWMOD®) that makes it possible to calculate a reference intensity ratio (RIR) for any clay mineral with a layer structure. This program makes it easy to calculate XRD peak ratios for the wide range of chemical varia- tions observed in natural samples. Furthermore, NEWMOD® gives peak intensities relative to non-clay minerals. Therefore, the calculated intensities can be used to check the clay mineral content, which is arrived at by proportioning the difference of 1 00% minus the non-clay min- eral content. The net result of these improvements is to minimize the errors associated with the mismatch between reference standards and minerals in coal and thus eliminate the need for an almost infinite number of natural standard samples. X-ray absorption error. The x-ray beam must pass through crystallite to be diffracted into the detector and recorded as a series of peaks. However, x-rays are absorbed in the crystallite in proportion to the atomic weight of its elements. Therefore, large grains of pyrite and marcasite will seem to be present at less than their true content because their diffracted x-rays are par- tially absorbed. In theory, highly precise and accurate analyses of these types of minerals require methods of sequential grinding to reduce errors to a minimum. In practice, it is possible to obtain adequate precision and accuracy with a single grinding in a McCrone® micronizer to produce 5-um to 1 0-um crystallites. Error in matching XRD peak intensities with mineral contents. Attempts to calculate the con- tent of all minerals directly from the XRD spectrum of a random bulk powder sample generally give erroneous results. This error was eliminated by applying a method that used chemical analysis of pyritic sulfur and XRD peak intensities from a random bulk pack to calculate the amount of pyrite, quartz, and calcite in LTA samples (Harvey et al. 1 983, Demir and Harvey 1 990). The bulk clay mineral fraction can then be calculated as 1 00% minus the non-clay min- eral fraction. By extension, the amount of calcite (confirmed to be present in the sample by XRD) can be calculated also from C0 2 or CaO analyses by traditional chemical methods, pro- viding that the concentration of other major Ca minerals in the samples is insignificant. For most coals, this calculation leaves only quartz among the major non-clay minerals in the sample to be determined solely with XRD. OBJECTIVES The objectives of this study were (1 ) to determine the mineralogical and chemical composition of the inorganic matter in marketed, conventionally cleaned Illinois coals from all operating mines and of CCRs from selected power plants; (2) to correlate the chemical composition with the mineralogical composition of the coals; (3) to investigate the fate of the inorganic constituents during combustion at the power plants; and (4) to determine the temporal variations in the mineralogical and chemical composition of marketed coals from selected mines. EXPERIMENTAL PROCEDURES Samples and Sample Preparation Marketed coal samples were collected from each of 35 coal preparation plants that operated in the different regions of the Illinois coal field (fig. 2) between 1 992 and 1 996. In 1 996, sets of fly ash, bottom ash, and feed coal samples were collected from a fluidized bed combustion (FBC) plant, two cyclone combustion (CYC) plants, and a pulverized coal combustion (PC) plant; a sample of the limestone used in the FBC plant was also collected. The coal sample from the FBC plant was a blend of coals from 1 7 different mines. Coal samples from the CYC and PC plants came from one of the 35 preparation plants that had been sampled four years earlier in 1 992. In addition, three other preparation plants among the 35 were sampled again four years after the first Figure 2 Coal regions of Illinois. Southeastern Illinois coal field is di- vided into Herrin (4H) and Springfield (4S) Coal seams. sampling. Results for these three samples and for the coal samples from the PC and CYC plants were compared with results from samples collected at the same plants four years earlier to study the temporal variations in the mineralogical and chemical composition of marketed Illinois coals. Details of location, collection, processing, and storage of the marketed coal samples from the coal preparation plants are described in Demir et al. (1 998). As for sampling at the power plants, the company staffs at the PC and CYC plants used automatic samplers to collect the samples as part of their routine sampling. For the FBC plant, we used a sampling shovel to collect 1 5 increments of the coal from the belt carrying coal to the boiler. To sample the FBC fly ash, the ash stream was directed to a sampling bag for about 30 sec through an opening at the particulate collection system. We sampled the FBC bottom ash and crushed limestone by using a sampling shovel to collect 15 to 20 increments of the materials from widely spaced locations of their stock piles. The power plant samples were weighed and described after a visual inspection (table 1 ). Representative splits of the coal and coarse-grained bottom ash samples were separated by riffling, splitting, and then grinding to -60 mesh (<250 urn) particle size in a Holmes mill. The fly ash samples were already 60 mesh in particle size and did not need to be ground. Small splits of the ground coal and bottom ash samples and of fly ash samples were prepared for chemical, mineralogical, and microscopic analyses by riffling and splitting. Table 1 Amounts and descriptions of samples of coal and CCRs from three types (FBC, CYC, PC) of power plants. Plant type Sample type Amount (kg) Sample description FBC CYC1 CYC2 PC coal 11.3 fly ash 8.2 bottom ash 10.9 limestone 11.3 coal 4.1 fly ash 3.6 bottom ash 5.4 coal 4.5 fly ash 3.2 bottom ash 5.4 coal 5.4 fly ash 3.6 bottom ash 6.8 <0.95-cm (3/8 in) particle size very fine particle size, light gray color <0.64-cm (<1/4 in) particle size, mostly yellowish and grayish particles, and small number of black particles crushed and off-white color <4.76-cm (-4 mesh) particle size fine particle size, dark gray color <0.95-cm (<3/8 in) particle size, black color, and high moisture content because of quenching in water <4.76-cm (-4 mesh) particle size fine particle size, dark gray color <0.95-cm (<3/8 in) particle size, black color, and high moisture content because of quenching in water <4.76-cm (-4 mesh) particle size fine particle size, light gray color <0.95-cm (<3/8 in) particle size, dark gray to black color, and high moisture content because of quenching in water Chemical Analysis of Coal and CCR Samples The samples of coals and CCRs were analyzed using the following methods: x-ray fluorescence spectrometry: Al, Ba, Ca, Fe, Ge, K, Mg, Mn, Mo, Na, P, Si, Sn, Sr, Ti energy-dispersive x-ray: Ba, Ge, Mo, Sn, Sr, Zr instrumental neutron activation: As, Br, Ce, Co, Cr, Cs, Dy, Eu, Ga, Hf, La, Lu, Mn, Rb, Sb, Sc, Se, Sm, Ta, Tb, Th, U, W, Yb, Zn atomic absorption spectrometry: Cd, Cu, Li, Ni, Pb, Zn, Hg (by cold vapor atomic absorption) optical emission spectrometry: Ag, B, Be, Ge, Mo, Pb, Sn, Tl, V pyrohydrolysis-ion chromatography: F inductively coupled plasma-mass spectrometry: Cd, Pb ASTM methods: moisture, ash, volatile matter, C, H, N, O, total S and S forms, CI, calorific value Note that some of the elements were determined using more than one method. In such cases, the average or the most reliable value was used. Chemical analyses of the samples were performed following strict quality assurance and quality control rules to ensure reproducibility and accuracy of the data. The quality assurance and control plan (ISGS 1 994) included standard operating proce- dures for sample preparation and for the analytical work. Blank, standard, and replicate samples were analyzed to ensure data reliability. Low-temperature Ashing of Coal Samples Generating the LTA of a coal sample was the first step for analysis of coal minerals. The LTAs were generated at about 1 50°C in an activated oxygen-plasma atmosphere that had been created by passing oxygen into a radiofrequency field (Gluskoter 1 965). The minerals in the LTA were essen- tially unaltered from their original state in the coal. In the standard procedure, 1 0-20 g of the sample were spread evenly to a depth of 3 mm over the bottoms of Pyrex boats (1 2 x 4 cm in size). The boats with samples were put in a vacuum desiccator for a minimum of 24 hr for drying. The dry samples were then placed in the asher, and the pressure in the asher was reduced to about 1 torr. The radiofrequency field was then introduced at 37.5 W per chamber, and oxygen gas was introduced at a rate of 20 cm 3 /min to produce the plasma. The ashing temperature was monitored occasionally with an infrared remote thermometer. The samples were stirred twice and weighed once each day. When the sample weight no longer changed, the ashing was complete. Water Leaching of LTA Samples Previous research indicated that the LTA samples contained some water-soluble, inorganic ele- ments that reduced the accuracy of the mineralogical analysis (Kruse et al. 1 994). To improve the accuracy of the mineralogical analyses, each LTA sample was washed with water to remove the water-soluble fraction. To accomplish the water extraction, each LTA sample was dispersed in demineralized water (1 g of LTA in 500 mL of water). After 12-16 hr, the LTA and water mixture was centrifuged, the clear supernatant was poured off, and the solid fraction was dried overnight at 70°C and weighed. Mineralogical Analysis of LTA Samples The water-extracted LTA samples were mineralogically analyzed by XRD. The data were then converted to whole-coal basis. For the XRD analysis, the LTA samples were micronized for 1 5 min with 1 mL of water buffered with a few drops of sodium dithionite in a McCrone® grinder. The samples were then transferred to 50-mL centrifuge tubes and centrifuged for 20 min at 2,000 rpm, which resulted in the separation of a clear supernatant. After the clear supernatant was discarded, the paste was mixed thoroughly, and some of it was smeared on glass XRD slides. The remaining paste was dried overnight at 70°C, mixed with a mortar and pestle, and then packed into an end- loading sample holder (a random bulk pack). The random bulk pack samples were x-rayed with a Scintag® diffractometer, and the peak areas were calculated with the DMS® software that is part of the instrument operating system. The smear slides were x-rayed after air-drying and after at least two days exposure to ethylene glycol, heating for 1 hr at 350°C, and then heating to 550°C, which produced the strong 14 A XRD peak for detection of small amounts of chlorite. The results of XRD analysis of bulk packs and smear slides were fairly comparable. The bulk pack XRD analysis was the preferred method for determination of non-clay minerals. The XRD analysis of smear slides was used for clay mineral analysis. The detection limit of calcite determination with XRD was 0.5 wt% or perhaps even greater. Therefore, the calcite content was calculated from the chemically determined CaO content of coal. This approach was judged to be reasonable because no other Ca minerals were detected in the samples by the XRD, and the amount of Ca associated with clay minerals and plagioclase was insignificant relative to that associated with calcite. The steps for calculating the mineralogical composition of the coal samples from the XRD spectra and chemical analyses were as follows: 1 . Deconvolute XRD patterns and transfer data electronically to a computer spreadsheet. 2. From the percent of insoluble LTA and chemically determined pyritic-S (S ) content of coal, %(pyrite + marcasite) in LTA = %S in LTA x 1 .871 py/ PV 3. From bulk pack XRD peak areas, calculate the percentages of quartz, K-feldspar, plagioclase, calcite, pyrite, and marcasite on a 100% basis. 4. Decimal non-clay minerals (DNC) in the LTA = (%[pyrite + marcasite] from step 2) -f (%[pyrite + marcasite] from step 3). 5. Percentage of each non-clay mineral in the LTA = (its percentage in the non-clay fraction) x (DNC). 6. From smear slide XRD peak areas, calculate the percentages of mixed-layered illite/smectite, illite, kaolinite, and chlorite on a 100% clay basis. 7. Percentage of each clay mineral in the LTA = (percentage in the clay fraction) x (1 -DNC). 8. Recalculate the percentage of calcite in LTA using the %LTA and chemically determined CaO contents of coal, replace this chemical calcite content for the XRD calcite content from step 5, and adjust the percentages of other minerals from steps 5 and 7 accordingly. 9. Percentage of each mineral in whole coal = (%LTA * 1 00) x (percentage in the LTA). Mineralogical Analysis of CCRs and Limestone The CCRs from the power plants contained large amounts of amorphous (glass) material and variable amounts of lime and anhydrite, which react with water. Therefore, a somewhat different procedure of mineralogical analysis was followed for the fly and bottom ashes as well as for the limestone sample: the water extraction step was omitted, 8 wt% dolomite was added to the sample as the internal standard, and the samples were micronized in propanol. Among a small number of available standards, the dolomite was the best because its peaks did not overlap with other mineral peaks, and 8 wt% dolomite produced enough XRD counts to ensure accuracy without diluting the other peaks too much. RESULTS AND DISCUSSION Mineralogical Composition of Marketed Coals On average, the most abundant minerals in the 35 marketed Illinois coals were kaolinite (x = 2.07%) and quartz (x = 2.06%)(table 2, fig. 3). K-feldspar was the least abundant mineral; its average concentration was 0.01 %. The clay mineral fraction generally was more abundant than the non-clay fraction in the 35 samples (fig. 4), and the clay mineral composition was slightly skewed towards kaolinite (fig. 5). The average clay content was 5.1 1% on whole coal basis or 52% on an LTA basis. This proportion of the clay fraction in the LTAs of the marketed Illinois coals is only slightly less than the 56% value for channel samples of Illinois coals reported in Harvey and Ruch (1 986), which suggests that conventional cleaning does not significantly alter the ratio of clay to non-clay minerals compared with that of standard channel samples. Table 2 Mineralogical composition of marketed basis.' linois coals and CCRs. All values are in weight percent on a dry weight Sample no. Region I/S I Kaol Chi Qtz Pyr Marc Calc K-f Plag C32773 1.67 0.79 1.50 0.07 2.24 1.42 0.24 0.71 0.05 0.07 C32774 0.43 0.53 0.93 0.05 1.26 2.35 0.26 0.27 <0.01 0.04 C32777 1.12 1.20 1.78 0.16 2.66 1.47 0.03 4.46 <0.01 0.08 C32778 1.79 1.23 2.31 0.23 1.82 0.91 <0.01 0.87 <0.01 0.06 C32782 0.61 1.10 1.69 0.06 3.62 2.00 0.06 1.27 <0.01 0.11 C32783 1.94 1.18 2.02 0.12 1.37 2.00 0.37 0.52 <0.01 0.10 C32785 1.83 1.21 1.77 0.09 2.16 1.54 0.29 0.64 0.03 0.05 C32797 1.87 1.06 1.89 0.13 1.87 1.64 0.06 0.77 <0.01 0.06 C32814 0.69 0.47 0.66 0.02 1.19 1.71 0.01 0.87 <0.01 0.07 C32779 2 1.87 0.77 1.34 0.21 2.25 1.68 0.09 0.77 <0.01 <0.01 C32794 2 1.09 1.20 2.34 0.17 2.36 2.02 0.12 0.79 <0.01 0.04 C32798 2 2.30 1.17 2.00 0.10 2.70 1.78 0.16 1.62 <0.01 0.06 C32800 2 2.70 1.17 2.23 0.20 2.17 1.41 0.18 0.82 <0.01 0.05 C32813 2 3.11 2.50 2.19 0.32 3.07 2.42 0.12 0.96 <0.01 0.05 C32815 2 1.65 1.03 1.88 0.13 2.95 2.17 0.11 1.52 <0.01 <0.01 C32784 3 1.81 1.00 2.34 0.18 1.41 1.26 0.09 0.20 <0.01 0.05 C32795 3 0.59 1.29 1.81 0.15 1.20 0.31 <0.01 0.18 <0.01 0.03 C32796 3 2.95 1.42 3.22 0.44 4.06 0.73 0.02 0.61 <0.01 0.11 C32799 3 1.06 2.65 3.21 0.44 1.42 0.34 0.01 1.89 <0.01 0.18 C32801 3 1.08 1.44 2.44 0.28 1.26 1.17 0.07 0.41 <0.01 0.07 C32802 3 1.39 1.73 2.09 0.60 1.32 1.42 0.10 0.71 <0.01 0.05 C32803 3 1.69 1.34 2.45 0.07 1.53 1.31 0.09 0.50 <0.01 0.05 C32661 4H 1.41 1.22 1.84 0.11 1.80 1.78 <0.01 0.34 <0.01 <0.01 C32664 4H 2.04 1.55 2.86 0.12 1.84 2.05 0.09 0.36 <0.01 0.06 C32665 4H 1.33 1.41 1.83 0.12 2.29 1.83 0.05 0.36 <0.01 0.03 C32771 4H 2.94 1.63 2.15 0.10 2.33 2.11 0.03 0.79 <0.01 0.05 C32776 4H 1.44 0.88 1.57 0.12 1.87 1.71 0.20 0.86 0.05 0.08 C32662 4S 1.48 1.20 1.64 0.19 1.15 1.01 0.06 0.23 <0.01 0.09 C32663 4S 1.31 1.18 2.80 0.19 1.77 1.30 0.53 0.29 <0.01 0.05 C32772 4S 1.10 1.19 2.19 0.18 2.24 1.50 0.19 0.80 0.02 0.09 C32775 4S 1.09 0.84 1.83 0.03 1.51 1.97 0.13 0.25 <0.01 0.14 C32780 4S 0.53 2.05 2.42 0.27 1.47 1.59 0.21 0.82 <0.01 0.06 C32781 4S 0.93 1.01 1.71 0.05 2.86 1.63 0.32 0.48 <0.01 0.07 C32793 4S 2.13 1.90 2.18 0.37 3.85 1.43 0.20 0.68 0.07 0.29 C35564 5 2.05 1.09 2.92 0.28 1.57 1.39 0.23 1.70 <0.01 0.06 Mean 2 all 1.57 1.29 2.07 0.18 2.06 1.56 0.13 0.84 0.01 0.07 SD 2 all 0.70 0.46 0.56 0.13 0.78 0.50 0.12 0.78 0.01 0.05 Power plant coals Sample no. Sample type l/S I Kaol Chi Qtz Pyr Marc Calc K-f Plag C35300 FBC feed 1.78 2.07 2.70 0.16 1.70 0.89 0.30 0.86 <0.01 0.00 C35301 CYC1 feed 0.46 1.47 2.14 0.04 3.03 1.12 <0.01 1.00 <0.01 0.13 C35302 CYC2 feed 0.60 1.54 2.13 0.30 2.79 1.06 0.05 1.05 <0.01 0.03 C35303 PC feed 0.35 2.32 3.47 0.19 1.61 1.27 0.07 1.00 <0.01 0.02 Power plant ashes Sample no. Sample type Mul Qtz Calc Hemat Mag i Anhyd Gyp s Lime Portl Glass C35304 FBC-fly ash ND 8.44 ND 3.03 0.69 17 ND 2.30 11 57 C35308 FBC-bottom ash ND 8.18 1.95 ND ND 16 2.03 40 13 19 C35305 CYC1-fly ash 1.87 7.57 0.75 2.96 4.03 ND 5.02 NC ND 78 C35309 CYC 1 -bottom ash ND 1.29 ND ND 0.44 ND 0.50 NC ND 98 C35306 CYC2-fly ash 2.29 6.83 1.11 2.16 3.79 0.60 ND ND ND 83 C35310 CYC2-bottom ash ND 1.27 ND ND 0.48 ND 0.29 NC ND 98 C35307 PC-fly ash 5.09 0.008 ND 2.07 4.74 0.71 ND NC ND 87 C35311 PC-bottom ash ND 0.97 0.37 1.97 4.95 ND ND NO ND 92 C35312 FBC - limestone3.4% illite. 1.5% kaolinite. 1.5% chlorite, 5% quartz, 86% calcite and 4.0% dolomite l/S, mixed layered illite/smectite; I, illite; Kaol, kaolinite; Chi, chlorite; Qtz, quartz; Pyr, pyrite; Marc, marcasite; Calc, calcite; K-f, K-feldspar; Plag, plagioclase; Mul, mullite; Hemat, hematite; Magn, magnetite; Anhyd, anhydrite, Gyps, gypsum; Portl, portlandite; Glass, amorphous material (calculated by difference); SD, standard deviation; H, Herrin Coal; S, Springfield Coal; ND, not detected For values below detection limits, one-half of the detection limits were used in statistical calculations of means and standard deviations. 2.5- £ 2- o "is 1-5- c CD O c o o CD D) CD i— 0) 1- 0.5- 27 38 45 36 71 | clay minerals ] non-clay minerals 32 92 92 71 100 Kaol l/S I Chi Qtz Pyr Marc Calc Plag K-f Figure 3 Average mineral content of 35 marketed coals. Numbers above bars are relative standard deviations ((standard deviation/mean) x 100). Keys: Kaol, kaolinite; l/S, illite/smectite; I, illite; Chi, chlorite; Qtz, quartz; Pyr, pyrite; Marc, marcasite; Calc, calcite; Plag, plagioclase; K-f, K-feldspar. 100% quartz + calcite 100% clay minerals 50 100% pyrite + marcasite Figure 4 Major mineral composition of mineral matter from 35 mar- keted Illinois coals. Relationships between Minerals and Minor and Trace Elements By using the mineralogical and chemical data (tables 2 and 3), statistical correlations were sought among the minerals and minor and trace elements in the marketed Illinois coals. The statistical analysis yielded many significant correlations (table 4). The square of the correlation coefficient (r 2 ) for some of the correlations was greater than 0.5 (figs. 6-1 0). Statistical correlations may or may not indicate that an element is located in the mineral with which it is positively correlated. For example, the correlation between quartz and Mg indicates that most of the quartz and the Mg- bearing mineral (illite/smectite) was incorporated into the coal by the same geological and geochemical conditions. The same interpretation can be made for the correlation between calcite and Zn. Calcite and the main Zn-bearing mineral (sphalerite) were incorporated into the coal by the 10 100% kaolinite 100% illite/smectite 50 100% illite + chlorite Figure 5 Clay mineral composition of the mineral matter from 35 marketed Illinois coals. 2.5 2.0 CD I.1-5H E oi in- o ' 0.5 0.0 y = 0.28x + 0.53 r* = 0.61 1 I ,dUU y = 185.00x + 273.60 m c 1 ,000 - r* = 0.53 ^ 800- E 600- cn 2 400- • 200- • • o- i i i 1 2 3 l/S, wt% 1 ,400 1,200-1 1,00 800 600 400 200 y= 152.69X + 348.30 r" = 0.52 1 1 1 - 2 3 l/S, wt% l/S, wt% Figure 6 Relationships of r 2 > 0.5 between elements and mixed-layered illite/smectite in marketed Illinois coals. same geochemical conditions, namely epigenetic precipitation; sphalerite was apparently present in the samples, but its concentration was too small to be detected by the XRD method. There is some evidence that F may be part of the illite and mixed-layered illite/smectite structure (Thomas et al. 1977). Manganese is known to replace Ca in the calcite structure, and, therefore, the Mn/calcite correlation was not surprising. Although the overall Hg-pyrite correlation is weak (fig. 1 1 ), there appear to be two populations, one above and one below the regression line, each with a different slope. This evidence suggests that Hg may be partitioned distinctly between pyrites of different origins (syngenetic vs. epigenetic). The modes of occurrence of most trace elements in coal are subject to multiple interpretations unless constricted by separating and directly analyzing individual minerals in coal. Combustion Behavior of Minerals in Coal Upon combustion, coal minerals (fig. 3) were converted to other minerals and amorphous phases that made up the CCRs (fig. 1 2). Among the minerals present in coal, only some of the quartz survived the combustion process; the rest of the quartz that was present in the coal was converted to silica glass. The clay minerals and feldspar were converted largely to aluminosilicate glass and, 11 £ 03 CD CD Ml _o CO (1) CO Li f- m m CO F CO c < (1) CO 0) < co c CO n CD CO fi) CO 3 s u. 0) u CO 01 CJ o (0 >> -i 0. o o~- E o CO o (A cu - ■n CO m O ai ^P O^ r o (1) 5? a> ° o T3 5 (0 ° a) c to Q. « i o <0 m o X) £ CD X !_L LL F ^o o in CD IB •fT <> CO U o co (/) > O 3? TO -Q ai -O ^c in CD m i F en .c u c/> m F < o CO o^ c o o CD 0) n i- CO t ^oiriocJcJcvJ^co^cviwojcNJcooocooJt^oco'^coui^co'cviTt^oo^'co'^tcoco CM t- 1-1—1- i- CO CO vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv ooooooooooooooooooooooooooooooooooo O i- •>- O i- t-Ot-i-t-OOi- OOOOOOOO"--- i- O i- O t- CO!Dl/)COtf)(DOUl>-NO)nCJ)NO)NmnCM'-^CDlO^'-aiOlJ)WaiCON001^ oi^ooi^oiOit^owr-^oir-Kinininr-qinNcDcoinaicqocviNocotDO) rir^c\i6wwriNwric\ic\i(\iT^c\idddd^wT^T^T^T^T^T^6T^T-'T^c\i^dci ^■cotoocMNMcooNconNcooinincococj^oiNnon'-cococoww^O) ■"in>-coo)4'-cDccic\icoitcotNNNONO)'-incocoNO)>-ir)i-no)nocon lod^codddd^d^d^dcMCMCMCMCMd^cMCMCMdco-^cM CO CO i- co T-0>COCOt-^CJCOI/lCJNC001IDCOCOU)-C\JCMl/)01CD(DCOlOlOl/)lflCOCDNCOO ^OCDOCDCNlCNlCDCOCDCOCvlCOi-M-^r^r^CDCOCvJlOCOOOOr^Or^i-LOr-i-CDLn dddddddd^dddrodmrodh^roi^KKKi^r^dror^r^dinr^''a : r~ '-T-cvjCMincOOCOCOifr-NOlOr-MOO)f(0'-(OCO'-'J*M-com"3-cMm -* ifi N 01'-CMO)CDCMinOOC , )^(MON'-(DCMCOO) a 2 CO CD CD CO CD m tft < < O) < CO CD CO 3 CO ° CO o >> a. o^ CO o c CO D) O ° CO CO o F O ^ O^ 3 5 ° « ° o T3 CD X LL ©^ CD CO o > o^ .Q 3 m .c (A < ° CD 3 CO O 2 ^p CO £■ i co n o id c g n o CO 0. _i i- * ■- >- e * in oi ■- m o oi >- OllllCONCOOCONIDN^CMO nnnnincMiocooiDcomcM ■•-cococomcococr) moooT-cnoooi^cncnmo aioiaicocMtfiNocMcomcO'- "-■-'-'-**Or CM CO CO V 00 CO ■* CM 1- NlDCOOinCKCDnniDCCINID oi*oinwNonosu)iD(ji toowi-qT-oisoiosow co" co" co" co" co' d cm" "*■" d d 08 o> co cor~r^cO'-cor^cD w cm cj cm n v in cm co co co i- v v v v v o o o o Ol N N O) ID CD ID N ci ci ci ci CM CM CM CM ^t n "t ir V V V V V V V o o o o m co cm ^r if oo r^ cx> CM CM CO CM ■= ■= CO CO CO SZ mOO ro>,>.m^ "•/-CM O "r ^ CJ >■"? ^ CM ° = o o o Voootooo-9 C0>>OCD>>Oc0>>-OcQ llOO0.llOOCLLlOOQ.ll Oi-CMCo^j-inior^oocDOi-CM oooooooooo-^-i-r: cococococococococococococo mminmininminminininin cococococococococococococo OOOOOOOOOOOOO 12 r^ot^coooor^ co o co q d d ^ d in in co co co r~- co in snomo u3Lnt/)LnLfi(Otf)c\j^tNNmcftT-c\ii^Ln v v v v v*-i-i- v v V i- i- i- v CM'-COCOCOCOCOCO'-COCO cOTrooM-M-oO"*-*'* CDCOC^COCDCOCMCOCOCOCO^rvJCOCOCO'TCMCO'JO"* *- co r^Oico cjcaaxomoeocoiDftiaxiDNflO i-cmoi-cjoioiconijiiii ^-t-O'-OOO'-'-OOOO'-OOOt-OOOO'-'-'-'-'-OOOOOO-^ oooooooooooooooooooooooooooooooooo CO CO CO m o co CO CO CM O) CD i- o d b o o o o o o OOOOOCJOOOOOO-3- OOtOCOCOIM'-T-(Di-'*^lOO) cocom-a-comm-a-co ^c\jP)roo)ifiiO(Cco(OincoN CMCOf^in^^CMlD^tOtDa) ^■"j-iritn'^fcnaj'-iiMooLncM cri m" ■** in o> in ■*" M* n in WWi-i-^OOOICOIMNNNt- oodicidicDC>dic>c>d>cid> cntoniDco^oosciiNinoNr- rimojoiwriri^di-Niciido TfiDiDiDCMniDioioNinnniDNcocO'-ino mcMOiD'-ioo CO CJ CO CO CO CO rt" r» co *- a> co co t^ •* •* ■* ** CD M; CM O CO CM CO O CO i^wsid^iniricoN o f in co NnroiDscooiiDNOOiioonMnsNsnon^oftniionMi^tMnmN o w N'-Ot\tN^OC\l^(OlDU)01N'tNtflroMD^OJCDO)miO^N^O)0)^lflO COCO mffioicoinsoioMDiDi-oiosin'-^mioNeoiooiiDiDMns'joiST-^ coin »-" -r-" -r-" i-" -r-" T-" T-" 1-" t-* C\f T-" CO* T-" «-" T-" CO C\J 1~" T-" T^ 1-" T^ C\T T-" l-~ ^-" T-" ^-" T-" !-" CO CM' *-" CTJ^-*^OOr-~T-(NTO-r-C0 OT-a)a)Mcoco^c»c\JO)^^ a>^f'd''3-r--T-cO'*a)T-c>oor^ oooooooooooooooooooooooooooooooooo OOOOOOOOOO V vo CO CM O) CD *- ^T 1^ CO r^ CD CM h- WlD^-COS^-Cfta)NT-C\lLnO) nnn^*^tt'-ra^m^iotinno)u)ifi^'^''t4in^oin'j"tt'*Noi m r- d d co mcoco^cNjO"- ddddojihinih CD co m- * i- O CM «- CM CM i- t 01(00 00 ON^^rocMWNOinnrjnwin^cjaicBCDO'-cDncO'-io^^'trvja) wric>iriwriwwnriri^coricj^rtc\iwririnririwriwn(\iw^'j CM CD co d N(DC0^00COO)NC0N(Dinin ocoo'Jcor-oooococnor^cocMCOTfr^'i-o-^-^ocoo'sfooCTiajOi-co 0)WCannul'-lDO)Wi-NN(I)NO(I>OlDOC»NNO^CMO)'JN'-00)lD^ jii-comofnoimconscomt-ostMinMDN^T-min^cj'-cJoinmN t-' t-" -r-" h~" m" in cm" r-" cm t-" in co" eg" m" m' d cm' n" co" d r-' d *-" cm* -3* cm" K cm" cm" co" co" cm" t-" in ,- CM ■>- ■t-r-i-t-r-T-'r-t-i-y-t-r- i- i- i- T- i- i- t- i- i- i- 1- 1- 1- 1- ODNN«001^010(00 cor-~ininiocMcoocDcncocoo co ■>- "^ ■* t^ M; cd ai in cm_ en cm_ t-_ cm" m" in" m" ■*" »-" ■*" co" d i"-" r-" rf is." r-i-i-i-^T-oiT-T-n^r- ococoooO'-miDn'-m^'OincoNnnNiDcoM-T-^r-^ncowNini-^o O)(ONKNC0T-T-l0a)O)C0C0l0C0(DU)CMNNC0O>lD(»C)NC0NlDC0NlDCja) h-coi-cocoooh-coinoom nioiDiDinnoiD't'-'-i- r- CM CM V V oooooooooooooooooooooooooooooooooo 0,-OOOOOOt-OOOO OOOOO'-O'-OOOOOi- CD O ■* CM co co in d C0OC0C0CMCDCDOinr^^CD^C0inOCMCMC0C0incDCMC0-r-;C0-t-;O codmdNdQ^coNNi/iosioVoitdcjcoVaiuiixiiDioc)'- CM i- O o en CO o o CM in 03 CM OOOOO-^t-i-Oi-'-t-o QQQQQQQQQClQQQ zzzzzzzzzzzzz ocomocooocoencoco*-o co cm co OOOOOO'-^O co h~-*cM r» -a- co co m **■ i-mooia)NO)coooc\j-iniq(DS(DOiooooneoinscjiDin**inso)i!!'-T co" in* cm" cm" co" co" co" co" CO co" co" CO cm" t"-' t-" CM CM »-" i- co" CO" t-" co" t-' CO" -r-" cm" CD O T- ■* d CD in CO ,_ CO CM *"" 1— o in o co '" in O] o o co in CO o CO o CO CO uiMOMONOinmwina)'- CMi--^T-r^CMCNJr~COi-CMCOT- CM CM -r- t- i- y- cDomLncocMincMcocMCMinm ^ o r- i- CD CO CD r- CD CD in O CMCO O) CO CO O) N in CMCM"d-'*incMcocO'*'*-'q-m ^dddw^id^ddddd T-CMNOOeONWOJCOOlCONCO noT-ooino~r>flMvnoo * o cm o o) to (O * n t- in w q co* "3* t" --C\l(0 CD XI h- > 10 c a % CD § o 0. > "a CD o CD CD CO => 3 £ E o F CO CD 3 ■o F en c o to £ 3 £3 E o u T3 e C CO en CO ■ o oo en co a> in ■•- ^- cm r^ in ^ co co r^-^-co wdNLow^roinSNomcd>-'-cdroco^^wwcoedri^c\icx)r : oiroinwi^w os^o^ininoioMMNcD^ocMni-eMnnmcocoiMoncMcononcocBif) ^fl- 1- T" fl- (Of T- *" CM cn^OTcoco^i^c3}eo^cor^CMmmr^co^r--.CMCocor'--coinin''3-i-"3'a>coco-3-cocM cMcocM^cM^ocor^^^^^co^^cor--.^ , ^-coTt-^tin^t-in^-(0^i-co^-cocoincD dooddooddooodcJoocr) - ' Dodddcidcidcidcidcicicid m mm mmmm in in in in in mmmmm cr>c^c^cocr>cr)^^c\imco^^mcAi^c>jcor^r^co'a-Tf'3-mmvf"a--^-Tfcr)mcr)mi- d d d d d NCOT-OlCOi-COCnO'-ONCliNCCOCOONOCNJ'-inCONOIXCIJCil CO O O CO CD fdfdr^frCliriNNNtlVfd'-df^rSN^Wf^drltiNNdd in r- o O ■* (') -r in tt *— CM (M vvvvvvvvvvvvvvvvv v v v v v V V V V V V aiooocr)'-r-^-r-;cx)0'-ccjmcx) fC\j^c\jr-^'-'odd'^ : 'r : cr)c\ic>j i-'-i-i-CMO-*'5fi-CMCMcr) nonifincona'tncjinionincttcminiiincBooioiNncoinoicooinnct mcomcDmmmmc\jmmcDc£)cx3cDm-^f-^cDcDmmmmmr^mmcDm'^mmcDco T- CO i- co CO i- V V V V r- O)00)ocof*uianoi0(00 nconcoNf>-inififCOci)tD in 4 in ■* oi i- r- ^_ (o o i- co n T-" to* d m" m-" -a-" co c\icOT-LnT-cocomcDcOTrNcO'-N^w i-^d^>-^-^,-^di-' J '--^cNJ-^'- J i- J CDin'JCD^NCDCO^CJ^nncMCMCO't in tlriN^T : ^dr ooooooooooooooooooooooooooooooooooo r- -^ cm *-c\j*-in cm cm ^ n >- cm ■■- ■■- -st- *- co i- co rOrT-^rrT-0^^i-T-ri-rrNrT-T-rrrrrrrFrrr>"»-N ddddddddddddddddddddddddddddddddddd coMoomcoo m mNcoinntM co a> i- cm ncooio co com «t COCMtOXtNSS COCOO)COlMO)C\JCOCO^(3)NO)CO'-0)nNNCO(3>0)NNnn d-r J dddddd^dddd-r J d' J d-r--^dddd-^d'-d-^ddddd'--^ coMcotocoor-c\i(0 0co>-sc\icooinc«coiOi-T-i-to^on( , ic»coO'-coin^ cO'-oiconcMCfinino^owN^'-ONcD^oiDcn^iDNtnN^^cD't^mco fflnwcO'finoNfflNinsrMncoinoocoi-coosNMOOi-innnoinioo o t-" r-." co" ^" d co" co" a>" cm" co" co" d m" m" co" m" •*" d d d d co" co" cm" d d d *-" cm" d d cm" co" W CMt-CMCMCMCMCMCM CMCNCNCMCOCM'-^'a-CMCMCMCMi-CMCMCOCMi-CMCMi-CMCMCOCO fflNioinffinffitoBBNo^f^fonioioinoios^sniooonifiT-N r : T^r : T^^^riT : ^cMT-ricMinwT^^w^T : ^^w^^r^^^i-^nc\ic\i'-ai cscoiDcDcoooni'-O'-in'fi-^cDcncoco'twcD^i-incoiciin^^nqsciis ^^^CMT-CMCMC^T^CMCMCMCMmCMCMT^dcMCMCMCMCM'dcMCMCMCMC^ NNrNin'-^iDC4tnco^siorfOfflffln«inio*nit)0'tNN't><;c\ir : d'-- J cM'dddd-r J dddddd'- J -r-i- J coddddddddi- J -^d»-'r-'- ; T^cM «)ONcoc\i'-'-»)ooco4CB'-n»)tif)'-'-"-o*ncO'-'-cno"-0'-cos mcomoomoomcomocor^o NnCli-i-OCMS OJcoinNco^'-co^cO'-cn^O'-ooir-aicDcncocfijisincocowT-co^romin co^*fflfeofflfNnsncocBNO)oin'-'-0)'-'-*0)*0'-T-coa)a)0>ooocococor-r^cocor-~r-~cocoa)co r^r^r^r^r^r s ^h-r^cor^r^r^cococor^-r v --r^r^cococococDcoh-r^cocor^r^r^r^r^m CNJCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCMin cocococococococoococomcoocococococococococococococococococococococo OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO CM o O o ^ o o o ro CO CM co in CM o o CO '- o -t CO CO o CD CMT-'-'-inOi-C0CMC0f--CDCM i-'-'-'-^nin^wcBocO'- ddddd-r-^i-^ddi-'dd mmcoo^^-M-comT-^rco CMCMCMOCOr^COOmCOO)CM COCMCOCMmCMCM'-m ONCDSCOCOlONNNincO'- fdddndcoioVdiiitf SrinulW^'-SMllNCOO) ffilDNNNOUl^NOinOO mOl^fin^l-COCOeMCOCD^ d M-" d in d r-.* a> co" r-" m-" d r-~" *-" CMCMCMCMC7)'-CO'*CO'*"*'-CM CM i- CM CM CM CM cr> 0> T " o ■st CD CM o CD 1^ O o o f- CM co omommmomoomr--.om'* SLNoioicd'-dcxicdS'-^ccinif) i-'-cO'-CM'-CMcncMin'-Tfcoco in CM m CM co CO CO -f in >* in CD CO mNoro^con* i- co oo m mcDcncnco-a-'-o-^-cOi-^m cM'-T- ; -f ; r^dddcMdd'^ : -r : M-SCOCDCOCDOICDCO'-OCMW 2 CO COCOCMCM"-r^ (J) CO CO CO CM o *ridn^docb*iri^diri t- co t--. ■* v v i- NCJCO^O^OIOICM CD-JNCMCOCM^tMD ^scoin^Lncocooj cocoNi-ioncos^ONinct i- t-^IONON'-MOCO t- CO noi^ocoinMnnO'-incD '-ncDCiin'-^CM'-coinni- r-ocoocococM'^'-nNaico ■r-" -r-" -r-" CO ■*" d T-" -r-" CO" CQ d O'-CMCO^incONCOOO'-M OOOOOOOOOO'-'-'- COCOCOCOCOCOCOCOCOCOCOCOCO mmmmmmmmmmmmm cococococococococococococo OOOOOOOOOOOOO a fe u> s en ID Cl> IT, CO o u CJ (/) CO C n 'I i D' CD n LL Li. 14 Table 4 Mineral-element correlations in marketed Illinois coals at a = 0.001 levels of significance, which correspond to r 2 > 0.30 for the number of samples used. Mineral-element pair Mineral-element pair lllite/smectite Cs Mg Ti Ta Sc Cr Rb Hf Ga Ba 0.62 0.53 0.52 0.49 0.48 0.47 0.46 0.42 0.41 0.36 lite + lllite/smectite Kaolinite Hf 0.58 Sc 0.55 Ba 0.53 Ti 0.53 Li 0.51 Ta 0.51 Th 0.46 Ce 0.38 La 0.36 Cs 0.32 Marcasite B 0.48 Br 0.37 Chlorite Quartz Calcite Mg F Ti Cs Rb Sc Th Ta Hf Ba Ga Ce La Ta Ti Hf Mg Ti Ta Mn Zn 0.66 0.63 0.63 0.60 0.59 0.56 0.56 0.55 0.49 0.47 0.45 0.37 0.34 0.34 0.33 0.31 0.53 0.32 0.31 0.94 0.54 to a lesser extent, to mullite. Pyrite and marcasite were converted to hematite, magnetite, and amorphous Fe oxides. Some of the calcite reacted with sulfur released from pyrite, marcasite, and organic matter and formed anhydrite, which later reacted with moisture and partially hydrated to gypsum. The rest of the calcite was converted to CaO (lime) and Ca(OH) 2 (portlandite); some of these minerals subsequently changed back to calcite through reaction with atmospheric C0 2 . At the FBC plant, large amounts of lime and portlandite were generated (mostly in the bottom ash) because of the limestone addition; subsequent reaction with atmospheric C0 2 most likely took place only on the exterior surfaces of these relatively coarse Ca minerals from the FBC ashes. The mineral conversion reactions apparently varied from one type of combustion unit to another, which resulted in distinct differences in the mineralogical as well as chemical compositions of the CCRs from the four coal-fired power plants. These differences have important implications for the effect of plant type on the HAPs emissions and on the suitability of the CCRs for use in cement, ceramic, and other commercial applications. Emission of HAPs from Power Plants If a HAP element is entirely retained in CCR, its emission into the atmosphere at coal-fired power plants is prevented. However, a portion of each HAP generally is mobilized and becomes airborne. The mobilized portions of the HAPs may be partially adsorbed on surfaces of hardware on the cool side (between the boiler and the stack) of a power plant and partially emitted into the atmosphere. Thus, the mobility values of HAPs would overestimate their atmospheric emissions, except in the case of highly volatile HAPs (F, Hg, Se). Therefore, it is difficult to obtain mass balance for trace elements from power plant samples, and the mobility values reported here should be used cau- tiously. 15 v = 153.21X + 114.83 1,000- 1^ = 0.66 ra 800- • "I / O) • Jg% E 600- *q$§* 2 400- 200- • i i i 4 6 l/S + I, wt% 300 250-1 c 200 g 150 "" 100- 50- y = 30.77x + 5.67 r 2 = 0.63 ~i 1 r 2 3 4 l/S +l, wt% 1,400 1,200- 1,000 800 600- 400 200 y= 125.09X + 221.53 r* = 0.63 • X* 2.5 2.0 -I 0.5- 0.0 y = 0.21x + 0.37 r 2 = 0.60 l/S + I, wt% 1 4 l/S + I, wt% 30 25 ra 20 H O) £ 15 °= 10- y = 2.73x + 3.23 r 2 = 0.59 4 6 l/S + I, wt% 3.5- y = 0.40x+ 1.12 r 2 = 0.56 % 3.0- • ^ |'2.5- E 2.0- • ^S>^ $ 1.5- ^w m 1.0- g 0.5- o.o- i i i 4 6 l/S + I, wt% 3.5- y = 0.34x + 0.52 r 2 = 0.56 3.0- • j£ 2.5- E 2.0- • 1 jE 1.5- 1.0- 0.5- • "" o.o- i i i l i 12 3 4 5 l/S + I, wt% 0.250 0.200 =* 0.150H E ^ 0.100 0.050 0.000 y = 0.02x + 0.05 r 2 = 0.55 ■"• • 2 4 l/S + I, wt% Figure 7 Relationships of r 2 > 0.5 between elements and mixed-layered illite/ smectite plus illite in marketed Illinois coals. 16 1.0 0.8 £ 0.6 E 0.2- 0.0 y = 0.16x + 0.12 _ r' = 0.58 # ■X- kaolinite, wt% 200 150- E100H CD 50- y = 26.42x - 11.92 r = 0.53 ■X 4.0 3.5 3.0 2.5 2.0 £ 1.5-1 1.0 0.5 0.0 y = 0.60x+ 1.03 l* = 0.55 kaolinite, wt% 1,400 1,200 1,000 H 800 ._- 600- F 400- 200 y = 173.49x + 219.44 f = 0.53 • X kaolinite, wt% kaolinite, wt% 50 40 H en =* 30 E 3 20 10 y = 8.11x 8.30 r* = 0.51 0.25 0.20 i ^£ 0.15 CD E ,« 0.10 H 0.05- 0.00 y = 0.03x + 0.04 r^ = 0.51 kaolinite, wt% — i 1 — 2 4 kaolinite, wt% Figure 8 Relationships of r 2 > 0.5 between elements and kaolinite in marketed Illinois coals. Figure 9 Relationship between quartz and Mg in marketed Illinois coals. 1,200 1,000 c? 800 H E 600 en 2 400 200 • • &• y= 167.18X + 212.96 r* = 0.53 quartz, wt% 17 The mobilities of 1 5 HAP elements from the four power plants were determined by using mass balance calculations (table 5). The mass balance of an element was calculated by comparing its mass in the feed (w F ) and in the CCRs (w CCRs ): Mass balance (%) = (w CCRs / w F ) x 100. The calculations of w CCRs and w F took into consideration the mass ratios of fly ash to bottom ash as well as the measured concentrations of elements and high-temperature ash in the samples. The fly ash to bottom ash ratios were 80/20 for the FBC plant, 25/75 for the CYC plants, and 75/25 for the PC plant. The feed for the FBC plant included not only coal but also the limestone added to the bed at a ratio of 75% coal to 25% limestone. The mass balances of the 1 5 elements were normalized to that of Al (by dividing them by Al mass balance and multiplying the result by 1 00) to minimize the effect of sampling and analytical errors on the mass balance computations. Aluminum was used for normalization because it is a refrac- tory element with a relatively large concentration in coal and is expected to be retained almost completely in the combustion ashes; no more than 5% (most of the time <1%) of Al was expected to escape the particulate collection systems with ultra-fine, airborne fly ash particles. Aluminum mass balances of approximately 1 00% for the CYC and PC units (Table 5) indicated that the mass balance calculations performed for these two types of plants were highly accurate. The Al mass balance for the FBC plant (76%) was somewhat small but still reasonable. The Al mass balance deficiency for the FBC plant could have resulted from variations in the composition of the feeds that would make it difficult to obtain a truly representative sample during a few hours of operation. The limestone and the coal used in the FBC furnace were blends of products from many different quarries and mines in Illinois. Therefore, future studies should consider collecting at least several 250 200- y = 41.43x + 4.11 r 2 = 0.94 500 400 =* 300 E c 200 100 y = 84.31 x + 18.63 r 2 = 0.54 2 3 4 calcite, wt% 1 calcite, wt% Figure 10 Calcite-Mn and calcite-Zn relationships in marketed Illinois coals. 0.30- 0.25- y = = 0.04x + 0.02 r = o.i3 « Figure 11 Relationship between pyrite and Hg in marketed Illinois en E di I 0.20^ 0.15- 0.10- 0.05- o.oo- • # • • • coals. i i i i 0.0 0.5 1.0 1.5 2.0 2.5 pyrite, wt% 18 40 30- 40 57 oS 20- FBC plant Fly ash Bottom ash 40- Mul Calc Qtz Magn Hemt Anhy Lime Portl Gyps Glass 87 92 PC plant 30- c g ']S 20 H O 8 10H Fly ash Bottom ash 40- 30- Mul Calc Qtz Magn Hemt Anhy Lime Portl Gyps Glass 78 98 '■5 20- 10- CYC plant Fly ash Bottom ash Mul Calc Qtz Magn Hemt Anhy Lime Portl Gyps Glass Figure 12 Mineralogical composition (from table 2) of CCRs from three types of plants (FBC, PC, CYC) in Illinois. The cyclone plant values were the averages of two plants. Key: Mul, mullite; Calc, calcite; Qtz, quartz; Magn, magnetite; Hemt, hematite; Anhyd, anhydrite; Portl, portlandite; Gyps, gypsum; Glass, amorphous phase. 19 "O ro CO CO « > c = CD < >_ CO n O Q. -JZ i_ O Zi o 5 s *" TO E £ o CO .Jz CO CD CO — t CO CO « O) CD O I? CO Q. E co = E T3 CO §§ co E QC c O - O _- CO 8 | c co — > E c 2 g 2 " s 3 c c o CD O F c -c O h- > H c = ^_ CO o > W LLI c DC CD —* E r CD c_ in f= "o E c CO Q} — > <"■§ o CO 73 c c O CO CO >* CO E J3 CO O CO H E -Q o r r U o O) fll in co zj § _o >. .a b ,? Q. E co CO T3 C CO c CO 03 1- o CO co it o iv i- 1- CD oo CM CM i- oo CO 00 LO oo CD ' r " IV <* C\j it cn CD o o CM CM - 1^ O) CM ,— d o 00 CO CD CO "3- ,_ o CM CD CM CD CD co d O 00 CD OS o CO CO CM CD CO CO q CO CD T- co CD "* CD "St q CO ^~ uo *™ *~ cd it d d T— it T_ ■>* o O ' I— ,— 1— LO - oo CD r ~ 1— d ' I— CM O (v 00 CO 1— O uo *t uo CO o iv CO ,_ r^ .^ o CM CO CM ,_ CD "* CM CO ,_ uo ■3-' V d V |v 00 CO o o CM CO <* CM N ■* CO LO ,_ o CM CM CM ' r_ CO oo co 1_ o CM |V CM CM CD |v d o ^ CO CO CM CO LO r-~ CD T- o CM oo q ^f T- CD CD CM ^ ■^t o 1 o d CD o o o o co 1 CM CD CO CO CM d d CO 1 co o CD o O IV C\J o O T— d <3- 00 ,_ oo co q r-. CM Tt CO rv LO CO CO 1 CM CM ■5t IV T _ CM CD CM CO CO ■^ -* it UO CO CD o CO O iv CO 00 CO CM UO 1^ LO CM CO d CM CO o CM ■* o o o o o ^f co d -3- d o d d A o O CD d d o iv 00 00 un CO <* uo UO LO CO d co d CO CD o 00 CM UO »~ o> oo d d CO o o CM LO CD O) d CM d i-~ co CO |v CD LO CO CM CD d d CD uo CM U0 LO CO 1 CO CD ^r d d IV rv CM CM CM U0 00 LO T— d 00 00 O CM LO CM UO CO LO 11 d r- LO |V CD CO CD CD 1 ^"" d CD CO 00 uo CD CM o CM CM CD LO LO 00 CM ■* LO LO CM CM CD r^ "St c\i <* CO CO IV C\J d d It CM d CM *t LO CO <* d d CO UO IV CO IV CM ,— d co d CM CM CM o CD O d o O o CO 00 CD UO UO CO CM cd uo 00 <* CM CO o CM CM cn <* h- CD CM CD CO |v CM ,_ ~ CM d d CO IV rv CM o O o ^ o o o CD o o CD o CM oo CM O 00 CM o *fr o CM OO oo o rv o U0 CVJ oo .^ o o LO ^r ,- 05 o o 00 CD o o o q o ,- CD d ,_ CD CD ifr o o it >* UO o CM o CO o o ,_ LO o CM CO co CO o o T ~ CD CD CD CD ,,— 1— co o co CD "sT CD CD CM q o ■* LO o LO o CM O o q .,_ 1— CD ■* co CO CD O O CM o LO 11 LO CD CM d CM CO -3- d o CD O CO o CM co CO CM O " o o o 03 OO CD o co CD CD co o oo CM O o o> q CM CO co CM OO q CM 00 IV "5t CD T _ CD T - CO ■>- d CD |v CM d d d rvl CO CM o o d 1_ d r^ t^- o O d 1_ d CO CD LO o O ,_ O CO CD CO CO CO CO -C co c w o S co Si CO H CO ro g CO CO ■c E CO to g o >> o o >, O CO o >, o E O ^ Dl = o -CZ CO co o 5= -Q CO co V =v -9 CM CM CM ^5 CO CO co to 2 o >, o o O O O O CD CD DO CD n x: E O > O O > > CD C) .c E O O O >■ >- >- C) CO CO E 6 6 6 CD o JZ co CO LL. LL LL LL C m ^5 co u O o o c rri ^p CO o o o o c m o o o CL 0- CL c_ CO cr- V, C) V, o ■>, u |v CO > O "* CO CM 10 £3 JD ■t— UO CD lb JO CM CD O JD >, iJ CO O f- ->, M— O O O i- i_ l_ o o o i_ i_ O O i- i_ i_ O CO i- 1_ C0 CO CO CO in — O o ro co co m — o o CO CO CO (D — o o CO CO CO CO r^ u LO LO LO LO (D .Q *♦" **~ LO UO LO CO JD **— M— LO LO LO CO JU 1 1 LO LO LO CO JJ ' CO CO CO CO CO O III III co CO CO CO O 111 III CO CO CO CO o LU LU CO CO CO CO o LU o o o o 2 ^ CC CC O O O ^ ^ DC DC O O O S s DC CL o o o ^ ^> CC 20 Table 6 Mobilities of HAPs during combustion. Mobility (%) Plant type Low (<25%) Moderate (25-50%) High (>50%) FBC As, Be, Co, Cr, Hg, Ni, P, Pb, Sb, Se, Th, U CYC As, Cd 2 , Cr, Ni, P 2 , Pb, Sb 2 , Th, U PC As 2 , Be, Cd, Cr, P, Pb, Sb 2 , Th Cd F, Mn Be, Co, Mn F, Hg, Se Co, Mn, Ni, U F Hg, Se 1 Average of two CYC plants. The two plants varied significantly with respect to the mobilities of some elements (see table 5). 2 Negative mobilities resulting from excess mass balance suggest the possibility of contamination of combustion ashes by erosion or corrosion of hardware (see text for more on the definition of mobility). sets of samples over a period of several months of operation from the FBC plant, and the average mass balance data on these samples should be used to smooth out the variance. The mobility of HAP elements from the four plants was calculated as follows: Mobility (%) = 100 - mass balance. According to this mobility relationship, if the amount of an element recovered from the CCRs accounted for 1 00% of the amount in the feed, then the element was assumed to be immobile. If the mass balance of an element was less than 100%, then the difference was considered to indicate the percentage of the element that entered and remained in the gas phase or condensed on the ultra-fine, airborne fly ash particles or on the power plant hardware. For convenience, the mobility values were divided into low, moderate, and high categories (table 6). Negative mobility values caused by the excess mass balance (101-1 67%) in some cases (table 5) probably resulted mainly from contamination of the CCRs by the erosion or corrosion of hardware in the combustion process. In particular, the excess mass balances observed for Cd for the CYC1 unit and for As, Cd, Pb, and Sb for the CYC2 unit (table 5) were too large to explain as sample variability or analytical error. Analysis of the CYC2 fly ash was repeated three times, and no significant variation was found between the three analyses. Although the negative mobility values were arbitrarily assumed to be in the low-mobility category, we recommend that periodic sampling of the CYC2 plant and analysis of the samples be carried out in the future to better understand the combustion behavior of As, Cd, Pb, and Sb at this unit. For the FBC plant, only the mobilities of F (85%) and Mn (58%) were large (table 6). The mobilities at the FBC plant were surprisingly low for the normally volatile elements Hg (18%) and Se (13%). Overall, the HAPs mobilities were lower at the FBC plant than at the other plants (table 5). Appar- ently, a relatively low combustion temperature (~900°C) or the chemical environment created by the addition of limestone resulted in generally lower HAPs mobilities. The FBC fly and bottom ashes naturally contained large amounts of anhydrite and lime-portlandite (table 3). Several investigators (Clarke and Sloss 1 992; Meij 1 993, 1 994a; Gullet and Ragnunathan 1 994; Bool and Helble 1 995; Querol et al. 1 995; DeVito and Bhagwat 1 997) reported that Ca-rich materials such as lime and limestone can capture substantial amounts of As, Hg, Sb, and Se during combustion. Suarez- Fernandez et al. (1 996), however, did not find any major differences between the combustion behavior of trace elements in a laboratory-scale FBC unit with or without the addition of limestone. This result suggests that low temperatures may influence the mobility of certain elements more than does limestone addition. 21 For the CYC and PC plants, mobilities of the volatile elements F (89-92%), Hg (75-90%), and Se (53-79%) were high (table 6), as expected. The mobility values of some elements for the CYC and PC plants were inconsistent with the volatility data summarized in figure 7 of Clarke and Sloss (1 992). According to Clarke and Sloss (1 992), As, Cd, Pb, and Sb are moderately volatile. However, our data indicated that all four elements had low mobilities at the PC plant, and Cd and Pb had low mobilities at the two CYC plants. Also, the Mn mobility was supposed to be low, which is inconsis- tent with the moderate Mn mobility calculated for the CYC and PC plants. Although Mn is not a volatile element at normal coal combustion temperatures, its mobility from the FBC plant was surprisingly large. Querol et al. (1995) reported that Mn has an affinity for Fe oxide in the CCRs. The Mn content of the feed coal from the CYC and PC plants was less than the Mn content of the feed (a blend of coal and limestone) from the FBC plant (table 5). Furthermore, CCRs from the CYC and PC plants contained more Fe oxide minerals (magnetite + hematite) than did the CCR from the FBC plant (table 3). These differences could be the reason Mn mobility was consider- ably greater at the FBC plant than at the CYC and PC plants. Studies reviewed by Clarke and Sloss (1 992) and Davidson and Clarke (1 996) indicated that among the 15 HAPs investigated, substantial portions of only F, Hg, and Se were mobilized in the gas phase and emitted into the atmosphere during coal combustion. The emission of other elements generally resulted from their enrichment in the submicron-size, fly ash particles that passed through the particulate control systems. Partitioning of HAPs between Fly Ash and Bottom Ash The atmospheric emission of elements from coal-fired power plants is directly related to the parti- tioning of elements among CCRs. For example, the enrichment of an element in the bottom ash would reduce its atmospheric emission. However, if an element is preferentially enriched in submi- cron-size, airborne, fly ash particles, its emission would be enhanced. The partitioning of elements in CCRs affects not only their atmospheric emissions but also the commercial value of the CCRs. To compare the distribution of HAP elements in fly and bottom ashes relative to their concentra- tions in coal, a relative enrichment factor (RE) was calculated for each HAP using the formula of Meij(1992): RE = (C CCR /C F ) x (%ash F /100). where C CCR and C F are the concentrations of an element in the CCR (fly ash or bottom ash) and feed, respectively, and %ash F is the percentage of ash in the feed. The feed refers to coal or, in the case of the FBC unit, a mixture of coal and limestone. Elements that are neither enriched nor depleted in the CCR should ideally have relative enrichment values of 1 . Elements with relative enrichment values of significantly greater or less than 1 are enriched or depleted, respectively, in the CCR. Based on the literature (Meij 1 992) and for convenience, the relative enrichment values in this study were divided into "normal," "enriched," and "depleted" categories (table 7). In most cases, the relative enrichment factor of the HAPs was greater for the fly ash than for the bottom ash samples from the four plants (table 5). The only reverse cases were Sb at the FBC plant, Mn at the CYC plants, and Co and Ni at the PC plant. The greater relative enrichment values for fly ash than for bottom ash probably resulted from the combination of two factors: (1 ) large portions of most HAP elements were volatilized during combustion and then, upon cooling, con- densed on the fly ash particles, and (2) some elements were enriched in finely disseminated coal minerals that ended up in the fly ash in greater proportions than they did in the coarse coal minerals. Some elements with large relative enrichment values could be recovered from the respective CCR and sold. For example, Se was highly enriched in the CYC1 fly ash (RE = 1 .7). The average con- centrations of Se in the earth's crust, shale, and soils are 0.05, 0.5, and 0.4 mg/kg, respectively (Clarke and Sloss 1992). As a result, the Se concentration in the CYC1 fly ash was 866, 87, and 1 08 times greater than the earth's crust, shale, and soil averages, respectively. Therefore, the 22 Table 7 Relative distribution ot HAPs among combustion ashes. Plant and Distribution ash type Normal (RE = 0.7-1.3) Depleted (RE < 0.7) Enriched (RE > 1.3) FBC-fly FBC-bottom Be. Cr. Hg. Ni. P. Pb, Sb, P Se, Th, U As, Cd, Co, F, Mn As, Be, Cd, Co, Cr, F, Hg, Mn, Ni, Pb, Se, Th, U none Sb CYC-fly CYC-bottom Be, Co, Hg, U Cr, Mn, Ni, P, Th, U F, Mn As, Be, Co, F Hg, Pb, Sb, Se As, Cr, Cd, Ni, P, Pb, Sb, Se, Th Cd PC-fly PC-bottom Be, Cd, Cr, Ni, R Th, U Be, Co, Cr, Ni, Th Co, F, Hg, Mn, Se As, Cd, F, Hg, Mn, P, Pb, Sb, Se, U As, Pb, Sb none 'Average of two CYC plants. CYC1 fly ash can be considered to be a potential source for commercial Se production. Rose et al. (1 979) reported that Se is produced industrially as a by-product of Cu refining; for an Arizona Cu ore deposit, for example, Se concentration was ~8 mg/kg.This value is much greater than the crustal average of 0.05 mg/kg Se, but similar in magnitude to the range of 7.4 to 43.3 mg/kg for Se found in the fly ashes from the three Illinois plants (table 3). In the past, relative enrichment values in the combustion products were used to assign the inor- ganic elements in coal to differing volatility classes (Clarke and Sloss 1 992, Meij 1 992, Davidson and Clarke 1996). However, such a task is often complicated by substantial variations in the combustion behavior of elements depending on the characteristics of coal, the type and operating conditions of power plants, and sampling and analytical errors. For example, some elements were enriched in the fly ashes from CYC or PC plants but not in the fly ash from the FBC plant (table 7). The only obvious trend common to all four power plants was that most HAPs were depleted or normally distributed in the bottom ashes (table 7). Although depletion of Al, a non-HAP conserva- tive element, is not expected in the CCRs, its depletion in the FBC bottom ash (RE = 0.3; table 5) is not surprising. The likely reason for the observed Al depletion is that the abundant and Al-rich clay minerals from the FBC feed coal ended up in the fly ash in greater proportions than the Al- deficient calcite that was the principal component of the limestone. This study did not examine the enrichment of the trace elements in particle-size fractions of the fly ash samples. Previous studies (Tumati and DeVito 1 991 , 1 993; Dale et al. 1 992; DeVito and Jackson 1 994; Helble 1 994; Meij 1 994b; Querol et al. 1 995; Cereda et al. 1 995; Suarez-Fernandez et al. 1 996) indicate that there is generally a negative correlation between ash particle size and the concentrations of As, Co, Cr, Hg, Mn, Ni,Sb, and Se. Temporal Variations in the Characteristics of Marketed Coals Marketed coals from four of the Illinois preparation plants were sampled twice with a 4-year interval between the two sampling times. The mineralogical and chemical characteristics of these as- shipped coals changed somewhat over the 4 years (table 8); some of these changes could partly be the result of the simple sample variation that can occur any time. There was no apparent systematic change that was common to all four coals. For example, the ash content increased for two of the coals and decreased for the other two. The temporal change in coal quality for some individual mines is not surprising because over time the production in a mine moves to different reserve blocks, commonly with somewhat differing coal qualities. And, in some cases, coal cleaning efficiency is altered over time to meet the specifications of coal buyers. For example, if the customer of a mine changes from a PC boiler without a scrubber to a PC boiler with a scrubber, a FBC unit, or gasification plant, the run-of-mine coal does not have to be cleaned as much for sulfur (pyrite) removal. The average coal quality for the entire basin, however, would not change significantly over a 4-year period. The coal C35300 from the FBC unit (Tables 3 and 5) was a blend of production from 1 7 different and mostly high-sulfur coal mines; thus, this sample probably represented a typical marketed coal from high-sulfur Illinois coal mines. 23 Table 8 Temporal variations in the mineralogical and chemical composition of marketed coals from four selected Illinois preparation plants. Sample 1 from each plant was collected in 1992 (Demir et al. 1994), and sample 2 from each plant was collected in 1996. All values are on a dry weight basis, except equilibrium moisture. Plant A Plant B Plant C Plant D Sample 1 Sample 2 Sample 1 Sample 2 Sample 1 Sample 2 Sample 1 Sample 2 Composition (C32773) (C35559) (C32774) (C35562) (C32782) (C35303)' (C32814) (C35561) Mineralogical (wt%) lllite/smectite 1.67 1.94 0.43 3.21 0.61 0.35 0.69 0.63 Mite 0.79 1.72 0.53 1.96 1.10 2.32 0.47 0.87 Kaolinite 1.50 2.85 0.93 1.60 1.69 3.47 0.66 1.33 Chlorite 0.07 0.14 0.05 0.20 0.06 0.19 0.02 0.05 Quartz 2.24 2.04 1.26 0.79 3.62 1.61 1.19 0.35 Pyrite 1.42 1.52 2.35 2.87 2.00 1.27 1.71 1.31 Marcasite 0.24 0.55 0.26 0.56 0.06 0.07 0.01 0.60 Calcite 0.71 0.91 0.27 0.21 1.27 1.00 0.87 0.45 K-feldspar 0.05 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Plagioclase 0.07 <0.01 0.04 0.11 0.11 0.02 0.07 <0.01 Ash 8.7 11.2 7.1 10.7 11.6 10.6 6.0 4.9 Volatile M 41.2 40.9 40.9 41.1 40.0 38.8 45.7 42.6 Fixed C 50.1 48.0 52.1 48.2 49.4 50.6 48.3 52.6 Pyritic S 0.97 1.20 1.46 1.87 1.25 0.79 1.09 1.11 Organic S 3.03 3.27 1.56 1.89 2.45 2.52 2.41 1.43 Sulfate S 0.14 0.19 0.56 0.19 0.20 0.69 0.18 0.51 Eq. moisture 13.9 12.1 13.4 12.2 14.3 14.7 11.3 13.9 Al 0.90 1.02 0.53 1.21 0.86 0.82 0.38 0.34 Si 2.10 2.53 1.13 1.94 2.74 2.55 1.00 0.67 Fe 1.19 1.41 2.11 2.25 1.50 1.55 1.26 1.67 Ca 0.29 0.36 0.11 0.09 0.51 0.40 0.35 0.18 K 0.16 0.20 0.091 0.27 0.16 0.15 0.075 0.06 Na 0.10 0.10 0.015 0.022 0.11 0.10 0.007 0.007 Chemical (mg/kg) As 1.3 1.6 20 26 2.4 2.2 6.4 46 B 135 163 116 98 106 181 83 161 Br 6.5 8.3 3.9 5 10 13.2 2 3.6 Ba 41 44 17 61 38 37 33 11 Be 1.0 <1 4.0 6.5 <1 1.0 2.0 3.0 Cd <0.3 1.0 <0.3 0.7 0.4 0.4 1.0 0.7 Ce 6.3 8.0 4.1 13.6 5.1 7.1 5.0 2.7 Co 2.3 3.4 3.1 7.5 1.6 3.5 2.5 11 Cr 12 3.4 5.8 14 14 17 5.7 5.7 Cs 0.88 1.1 0.75 1.8 0.68 1.0 0.92 0.80 Cu 7.4 8.2 13.6 18.2 6.2 5.0 8.8 20.5 Eu 0.16 0.21 0.32 0.52 0.1 0.10 0.3 0.35 F 90 141 68 133 78 63 63 50 Ga 3.2 2.9 4.0 5.3 2.1 2.4 2.2 2.7 Ge <5 <5 27 61 6 7 16 125 Hf 0.36 0.5 0.32 0.6 0.4 0.6 0.17 0.2 Hg 0.13 0.07 0.22 0.15 0.07 0.08 0.07 0.13 La 3.5 4.3 2.2 7.1 3.1 2.9 1.7 1.2 Li 3.9 4.3 5.8 14.1 2.8 3.3 0.9 1.2 Lu 0.1 0.09 0.11 0.17 0.07 0.10 0.12 0.14 Mg 482 603 241 663 543 543 241 121 Mn 39 42 18 24 55 43 30 20 Mo <6 6.5 <5 <5 11 8.0 15 <3 Ni 11 11 15 22 7 11 12 20 P 44 87 87 87 87 44 30 87 Rb 11 14 5 19 8.2 12 5.1 4.3 Pb <6 5.4 102 169 <6 3.2 23 184 Sb 0.2 0.3 1.2 2.3 0.5 0.6 1.9 5.1 Sc 1.9 2.0 1.9 3.4 1.6 1.9 1.1 1.5 Se 1.9 2.7 1.2 1.6 1.9 2.9 1.3 1.9 Sm 0.65 1.0 1.2 2.4 0.49 0.70 1.0 1.3 Sn <5 <5 <5 <5 <5.5 <5 <5 <5 Sr 28 33 17 33 25 26 17 10 Ta 0.1 0.1 0.07 0.1 0.1 0.1 0.05 <0.1 Tb 0.12 0.11 0.25 0.44 0.07 0.10 0.29 0.29 Th 1.2 1.4 0.8 2.1 1.1 1.4 0.6 0.6 Tl <1 <1 <1 <1 1 <1 <1 <1 Ti 539 599 300 599 539 480 240 240 U 1.7 1.3 <0.8 1.0 1.3 2.3 8.0 0.8 V 16 36 11 26 23 41 16 16 w 0.35 0.50 0.2 0.3 0.35 0.4 0.25 0.10 Yb 0.29 0.40 0.61 1.1 0.26 0.5 0.76 0.73 Zn 33 99 71 287 42 34 68 179 Zr 20 11 19 21 23 9 11 9 Two other newly collected coals (tables 3 and 5, CYC 1 and CYC 2 feed coals) came from the same mine as the sample C35303. 24 SUMMARY AND CONCLUSIONS The mineralogical and chemical composition of (1 ) marketed Illinois coals from 35 mines and (2) feed coal, fly ash, and bottom ash from three different types of power generating units (PC, CYC, and FBC) burning Illinois coals were investigated. This investigation was the first time the nature and combustion behavior of mineral matter in marketed Illinois coals were determined in such a systematic and detailed manner. The general findings were as follows: • The mean clay mineral content of the marketed Illinois coals was about 52% of the total mineral matter present in these coals. The mean values for abundances of the individual minerals in the coals (with standard deviations shown as ± values) were kaolinite, 2.07 ± 0.56%; quartz, 2.06 ± 0.78%; mixed-layered illite/smectite, 1 .57 ± 0.70%; pyrite, 1 .56 ± 0.50%; illite, 1 .29 ± 0.46%; calcite, 0.84 ± 78%; chlorite, 0.18 ± 0.13%; marcasite, 0.13 ± 0.12%; plagioclase, 0.07 ± 0.05%; and K-feldspar, 0.01 ± 0.01%. Quartz partially survived the combustion process and was par- tially converted to silica glass at the power plants. The clay minerals and feldspars were con- verted to aluminosilicate glass and to mullite. Pyrite and marcasite were converted to hematite, magnetite, and amorphous Fe oxides. Calcite was converted to lime, portlandite, anhydrite, and gypsum. • Concentrations of some minor and trace elements increased with increased concentrations of certain minerals in the 35 coals. Among the 15 HAP elements (As, Be, Cd, Co, Cr, F, Hg, Mn, Ni, P Pb, Sb, Se, Th, U) in the coals, the concentrations of Cr, F, and Th were strongly and positively correlated with the content of mixed-layered illite/smectite or the combination of mixed- layered illite/smectite and illite. Strong positive correlations existed also between Th and kaolinite contents and between Mn and calcite contents. There were weak positive correlations between Co and chlorite, Hg and marcasite, and Hg and pyrite. These relationships between minerals and elements are important considerations for designing methods for pre- and post-combustion con- trol of HAP emissions from coal-fired power plants. • Upon combustion, a portion of each HAP element is mobilized and becomes airborne. The mobilized portions of an element could partially be adsorbed on surfaces of hardware on the cool side of a power plant instead of being emitted into the atmosphere. In that case, the atmospheric emission of HAPs, except highly volatile ones (F, Hg, Se), could be less than the emission values reported here. Therefore, in this study, the term "mobility" instead of "emission" was used to express the portion of an element that was not retained in CCRs.The mobilities of 1 5 HAPs from the FBC plant were generally less than the mobilities from the other plants. Relatively large mobilities (more than 50% of the original amount in coal) were observed for F and Mn at the FBC plant and for F, Hg, and Se at the CYC and PC plants. • Combustion mobilities of trace elements are controlled by their volatility and affinity for various coal combustion phases. The elements investigated in this study occurred in greater concentra- tions in the fly ash than in the bottom ash in most cases, which is generally consistent with the literature data. The relative enrichment of the elements in the fly ash most likely resulted from the combination of (1 ) the partial volatilization of the elements from the bottom ash and their subsequent condensation on the fly ash particles upon cooling and (2) preferential enrichment of some elements in finely disseminated coal minerals that ended up largely in the fly ash. Some CCRs contained high concentrations of certain elements that could be recovered and sold. • There were differences in the mineralogical and chemical compositions of two sets of marketed coal samples taken from four Illinois preparation plants at a 4-year interval. Some of the differ- ences could result from expected sample variations independent of when the samples were col- lected. However, it is likely that the observed differences largely resulted from changes in char- acteristics of in-place coal as mining moved to different reserve blocks over time and from changes in coal cleaning in response to market demands. Therefore, historical coal quality data on clean coal products from a given mine cannot be used to accurately determine the quality of the current production from the same mine. 25 REFERENCES ACAA, 1 999, Coal combustion product production and use for 1 998: Alexandria, Virginia, American Coal Ash Association. ASTM, 1995, Annual book of ASTM standards, section 5, volume 0.505 — Gaseous fuels; coal and coke: Philadelphia, Pennsylvania, ASTM. Barbieri, L, I. Lancelotti, T. Manfredini, I. Queralt, J. M. Rincon, and M. Romero, 1999, Design, obtainment and properties of glasses and glass-ceramics from coal fly ash: Fuel, v. 78, p. 271-278. Bool, L.E., and J.J. Helble, 1 995, A laboratory study of the partitioning of trace elements during pulverized coal combustion: Energy and Fuels, v. 9, p. 880-887. Cereda, E.M.,G.M. Braga, M. Pedretti, G.W. Grime, and A. Baldacci, 1995, Nuclear microscopy for the study of coal combustion related phenomena: Nuclear Instruments and Methods in Physics Research B, v. B99, no. 1/4, p. 414-418. Clarke, L.B., 1 992, Application for coal-use residues: London, International Energy Agency (IEA) Coal Research IEACR/50. Clarke, L.B., and L.L. Sloss, 1992, Trace elements — Emission from coal combustion and gasifica- tion: London, IEA Coal Research IEACR/49. Dale, L.S., S.A. Lawrensic, and J.F. Chapman, 1 992, Mineralogical residence of trace elements in coal — Environmental implications of combustion in power plants, CSIRO investigation report CET/IR058: North Ryde, New South Wales, Australia, CSIRO, 79 p. Davidson, R.M., and L.B. Clarke, 1996, Trace elements in coal: London, IEA Coal Research, IEAPER/21. DeBarr, J. A., D.M. Rapp, M. Rostam-Abadi, J.M. Lytle, and M.L. Rood, 1996, Valuable products from utility fly ash: Cartersville, Illinois, Illinois Clean Coal Institute Final Report. Demir, I., and R.D. Harvey, 1 990, Abundance and origin of major minerals in the Herrin Coal, Illinois Basin, in R. W. Bryers and K. S. Vorres, eds., Mineral matter and ash deposition from coal: New York, Engineering Foundation, p. 13-22. Demir, I., R.D. Richard, R.R. Ruch, H.H. Damberger, C. Chaven, J.D. Steele, and W.T. Frankie, 1 994, Characterization of available (marketed) Illinois coals from Illinois mines: Illinois State Geological Survey Open File Series 1 994-2. Demir, I., R.R. Ruch, H.H. Damberger, R.D. Harvey, J.D. Steele, and K.K. Ho, 1998, Environmen- tally critical elements in channel and cleaned samples of Illinois coals: Fuel, v. 77, p. 95-107. DeVito, M., and S.B. Bhagwat, 1 997, Correlate coal/scrubber parameters with Hg removal and Hg species in flue gas: Paper presented at the 1 5th Annual Contractors' Meeting of Illinois Clean Coal Institute, July 29-30, 1 997, Champaign, Illinois. DeVito, M.S., and B.L. Jackson, 1994, Trace element partitioning and emissions in coal-fired utility systems: Paper presented at the 87th Annual Meeting and Exhibition of Air and Waste Man- agement Association, June 1 9-24, 1 994, Cincinnati, Ohio. Dhir, R.K., and M.R. Jones, 1999, Development of chlorite-resisting concrete using fly ash: Fuel, v. 78, p. 137-142. Dreher, G.B., W.R. Roy, and J.D. Steele, 1 996. Laboratory studies on the codisposal of fluidized- bed combustion residue and coal slurry solid: Illinois State Geological Survey Environmental Geology 150. Galbreath, K., C. Zygarlicke, G. Casuccio, T. Moore, P. Gottlieb, N. Agron-Olshina, G. Huffman, A. Shah, N. Yang, J. Vleeskens, and G. Hamburg, 1 996, Collaborative study of quantitative coal mineral analysis using computer-controlled scanning electron microscopy: Fuel, v. 75, p. 424-430. Gluskoter, H.J., 1 965, Electronic low temperature ashing of bituminous coal: Fuel, v. 44, p. 285-291 . 26 Gluskoter, H.J., R.R. Ruch, W.G. Miller, R.A. Cahill, G.B. Dreher, and J.K. Khun, 1977, Trace elements in coal: Illinois State Geological Survey Circular 499, 154 p. Gullet, B.K., and K. Ragnunathan, 1 994, Reduction of coal-based metal emissions by furnace sorbent injection: Energy and Fuels, v. 8, p. 1068-1076. Harvey, R.D., R.A. Cahill, C.-L. Chou, and J.D. Steele, 1 983, Mineral matter and trace elements in the Herrin and Springfield Coals, Illinois Basin Coal Field: Illinois State Geological Survey Contract/Grant Report 1 983-4. Harvey, R.D., and R.R. Ruch, 1 986, Mineral matter in Illinois and other coals, in K. S. Vorres, ed., Mineral matter and ash in coal, American Chemical Society Symposium Series 301 : Washing- ton, D.C., American Chemical Society, p. 10-40. Helble, J.J., 1 994, Trace element behavior during coal combustion: Results of a laboratory study: Fuel Processing Technology, v. 39, p. 1 59-1 72. Hughes, R.E., P.J. DeMaris, G.B. Dreher, D.M. Moore, and M. Rostam-Abadi, 1996, Brick manu- facture with fly ash from Illinois coals: Carterville, Illinois, Illinois Clean Coal Institute Final Technical Report. Hughes, R.E., D.M. Moore, and H.D. Glass, 1994, Qualitative and quantitative analysis of clay minerals in soils, in J .E. Amonette and L.W. Zelazny, eds., Quantitative methods in soil mineralogy: Madison, Wisconsin, Soil Science Society of America Miscellaneous Publication, p. 330-359. Hughes, R.E., and R.L. Warren, 1 989, Evaluation of the economic usefulness of earth materials by x-ray diffraction, in R.E. Hughes and J.C. Bradbury, eds., Proceedings of the 23rd Forum on the Geology of Industrial Minerals: Illinois State Geological Survey Industrial Minerals Notes 102, p. 47-57. ISGS, 1 994, QA/QC plan for the analytical geochemistry section: Illinois State Geological Survey internal document. Kruse, C.W., C. Chaven, R.E. Hughes, R.R. Ruch, and J.M. Lytle, 1994. Illinois Basin coal sample program: Carterville, Illinois, Illinois Clean Coal Institute Final Technical Report. Meij, R., 1992, A mass balance study of trace elements in a coal-fired power plant with a wet fgd facility, in G. Vourvopolis, ed., Elemental analysis of coal and its by-products, River Edge, New Jersey, World Scientific Publishing Co., p. 299-31 8. Meij, R., 1 993, The fate of trace elements at coal-fired power plants, in W. Chow and L. Lewin, eds., Proceedings of 2nd International Conference on Managing Hazardous Air Pollutants: Palo Alto, California, Electric Power Research Institute EPRITR-1 04295, p. V83-V105. Meij, R., 1 994a, Distribution of trace species in power plant streams: A European perspective, in Proceedings of 56th Annual Meeting of American Power Conference, April 25-27, 1 994, Chi- cago, Illinois, v. 56-1 , p. 458-463. Meij, R., 1 994b, Trace element behavior in coal-fired power plants. Fuel Processing Technology, v. 39, p. 199-217. Moore, D.M., and R.C. Reynolds Jr., 1 997, X-ray diffraction and the identification and analysis of clay minerals: New York, Oxford University Press. Moore, T., 1 994, Hazardous air pollutants: measuring in micrograms: EPRI Journal, January/ February, p. 7-15. National Research Council, 1 980, Trace element geochemistry of coal resource development related to environmental quality and health: Washington, D.C., National Academy of Science, 1 53 p. Niewiadomski, M., J. Hupka, R. Bokotko, and J.D. Miller, 1999, Recovery of coke fines from fly ash by air sparged hydrocyclone flotation: Fuel, v. 78, p. 161-168. Querol, X., J.L. Femandez-Turiel, and A. Lopez-Soler, 1995, Trace elements in coal and their behavior during combustion in a large power station: Fuel, v. 74, p. 331-343. 27 Rao, C.P., and H.J. Gluskoter, 1973, Occurrence and distribution of minerals in Illinois coals: Illinois State Geological Survey Circular 476. Reynolds, R.C., Jr., 1 985, NEWMOD®, a computer program for the calculation of one-dimensional diffraction patterns of mixed-layered clays: Hanover, New Hampshire, published by R.C. Reynolds, Jr. Reynolds, R.C, Jr., 1989, Principles and techniques of quantitative analysis of clay minerals by x- ray powder diffraction, in D. R. Pevear and F. A. Mumpton, eds., Quantitative mineral analysis of clays: Evergreen, Colorado, Clay Minerals Society, p. 4-36. Rose, A.W., H.E. Hawkes, and J.S. Webb, 1979, Geochemistry in mineral exploration (2nd ed.): New York, Academic Press. Roy, W.R., G.B. Dreher, J.D. Steele, R.G.Darmody, D.Tungate, W.E. Giles, and S.C. Phifer, 1997, Direct revegetation of coal slurry after amendment with FBC residues: Carterville, Illinois, Illinois Clean Coal Institute Final Technical Report. Russell, S.J., and S.M. Rimmer, 1979, Analysis of mineral matter in coal, coal gasification ash, and coal liquefaction residues by scanning electron microscopy and x-ray diffraction, in C. Kan Jr., ed., Analytical methods for coal and coal products, v. 3: New York, Academic Press, p. 133-162. Slaughter, M., 1989, Quantitative determination of clay and other minerals in rocks, in D. L. Bish and J. E. Post, eds., Quantitative mineral analysis of clays, Washington, D.C., Mineralogical Society of America, p. 183-216. Sloan, J.J., R.H. Dowdy, M.S. Dolan, and G.W. Rehm, 1 999, Plant and soil response to field- applied flue gas desulfurization residue: Fuel, v. 78, p. 169-174. Snyder, R.L., and D.L. Bish, 1989, Quantitative analysis, in D.L. Bish and J.E. Post, eds., Modern powder diffraction, Reviews in mineralogy, v. 20: Washington, D.C., Mineralogical Society of America, p. 101-144. Suarez-Fernandez, G., X. Querol, J.L. Femandez-Turiel, A.B. Benito-Fuertes, and M.R. Martinez- Tarazona, 1 996, The behavior of elements in fluidized bed combustion: American Chemical Society Fuel Chemistry Preprints, v. 41 , p. 796-800. Swaine, D.J., 1990, Trace elements in coal: London, Butterworth, 276 p. Thomas, J. Jr., H.D. Glass, W.A. White, and R.M. Trandel, 1977, Fluoride content of clay minerals and argillaceous earth materials: Clays and Clay Minerals, v. 25, p. 278-284. Tumati, P.R., and M.S. DeVito, 1 991 , Retention of condensed/solid phase trace elements in an electrostatic precipitator, in Managing hazardous air pollutants — State of the art: Palo Alto, California, Electric Power Research Institute. Tumati, P.R., and M.S. DeVito, 1993, Trace element emissions from coal combustion — A compari- son of baghouse and especially collection efficiency, in Proceedings of 3rd International Con- ference on Effects of Coal Quality on Power Plants: Palo Alto, California, Electic Power Research Institute EPRI TR-1 -2280, p. 335-348. U.S. Public Law 1 01 -549, 1 990, Clean Air Act Amendments, Title 3, 1 04 Statute 2531-2535. USEPA, 1998, Study of hazardous air pollutant emissions from electric utility steam generating units, Final report to Congress: v. 1 and 2, EPA-453/R-98-004b. Ward, C.R., 1 977, Mineral matter in the Springfield-Harrisburg (No. 5) Coal Member in the Illinois Basin: Illinois State Geological Survey Circular 498, 35 p. Wesnor, J.D., 1 993, ABB's investigation into the utility air toxic problems, Paper presented at EPRI/EPA S0 2 Conference, Boston, Massachusetts. Wirtz, G.P., and J.M. Bukowski, 1996, Extruded honeycomb construction materials from fly ash: Carterville, Illinois, Illinois Clean Coal Institute Final Report. 28