:G150 /*f. &S: (La T Laboratory Studies on the Codisposal of Fluidized-Bed Combustion Residue and Coal Slurry Solid Gary B. Dreher, William R. Roy, and John D. Steele M ^ 1 5 1997 ■v/A^v •4« ' L *-•-> nvtY Environmental Geology 150 1996 Department of Natural Resources ILLINOIS STATE GEOLOGICAL SURVEY Laboratory Studies on the Codisposal of Fluidized-Bed Combustion Residue and Coal Slurry Solid Gary B. Dreher, William R. Roy, and John D. Steele MAY 1 5 1997 Environmental Geology 150 1996 ILLINOIS STATE GEOLOGICAL SURVEY William W. Shilts, Chief Natural Resources Building 615 East Peabody Drive Champaign, Illinois 61820-6964 (217)333-4747 GLOSSARY OF ABBREVIATIONS pm micrometer ISGS Illinois State Geological Survey AAS atomic absorption spectrometry K potassium Ag silver K + potassium ion aglime agricultural limestone L liter Al aluminum mg milligram As arsenic Mg magnesium B boron Mg 2+ magnesium ion Ba barium MgC0 3 magnesite Ca calcium MgO periclase Ca 2+ calcium ion MgS0 4 ° magnesium sulfate ion pair CaC03 calcite Mn manganese Ca(OH) 2 portlandite Mn 2+ manganese ion CaMg(C0 3 ) 2 dolomite Mo molybdenum CaO lime Na sodium CaS0 4 ° calcium sulfate ion pair Na + sodium ion CaS0 4 anhydrite Na 2 S0 4 ° sodium sulfate ion pair CaS0 4 -2H 2 gypsum Ni nickel CaS0 4 « 1 / 2 H 2 bassanite 2 oxygen gas Cd cadmium OEP optical emission (photographic) CI chlorine spectroscopy cr chloride ion Org-C organic carbon Co cobalt Org-S organic sulfur co 3 2 " carbonate ion Pb lead co 2 carbon dioxide Pyr-S pyritic sulfur Cr chromium S sulfur CSS coal slurry solid Se selenium Cu copper SEM-EDS scanning electron microscopy Eh oxidation-reduction potential with energy-dispersive X-ray F fluorine spectrometric detection F fluoride ion Si silicon FBC fluidized-bed combustion Si0 2 quartz Fe iron S0 4 2 ~ sulfate ion FeS 2 pyrite S0 2 sulfur dioxide GF-AAS graphite furnace atomic Sul-S sulfatic sulfur absorption spectrometry Tot-C total carbon H 2 S0 4 sulfuric acid Tot-S total sulfur Hg mercury- V vanadium IC ion chromatography XRD X-ray diffraction spectrometry INAA instrumental neutron activation XRF X-ray fluorescence spectrometry analysis Zn zinc In-C inorganic carbon Zr zirconium ILLINOIS DEPARTMENT Of NATURAL RESOURCES Printed by authority of the State of Illinois/1996/1,000 © Printed with soybean ink on recycled paper CONTENTS ABSTRACT 1 INTRODUCTION 1 MATERIALS AND METHODS 2 Characterization of the Unleached Solids 3 Preparation of Mixtures 5 Wet-Dry Leaching Experiments 5 Batch Extraction Experiments 5 Analysis of Leachates and Extracts 6 RESULTS AND DISCUSSION 6 Physical, Chemical, and Mineralogical Characteristics of FBC Residues and CSS 6 Particle size 6 Chemical composition 6 Mineralogical composition 6 Effects of Leaching 1 1 pH of Leachates and Extracts 14 Major Solutes in Leachates and Extracts 16 SUMMARY 22 ACKNOWLEDGMENTS 24 REFERENCES 24 APPENDIX 26 FIGURES 1 Particle-size distribution of FBC residues, aglime, and CSS materials 7 2 X-ray diffraction spectra of FBC-1 and FBC-4 10 3 X-ray diffraction spectra of FBC-1 and CSS-2 before wet-dry leaching and CSS-2/FBC-1 after 15 pore volumes of water were leached through the mixture 14 4 pH in leachates and extracts from unmixed FBC residues and CSS 15 5 pH of leachates and extracts from mixtures of FBC residues and CSS-1 17 6 pH of leachates and extracts from mixtures of FBC residues and CSS-2 17 7 Calcium concentration in leachates and extracts from mixtures of FBC residues and CSS-1 18 8 Calcium concentration in leachates and extracts from mixtures of FBC residues and CSS-2 18 9 Sulfate concentration in leachates and extracts from mixtures of FBC residues and CSS-1 19 10 Sulfate concentration in leachates and extracts from mixtures of FBC residues and CSS-2 19 1 1 Molar concentrations of calcium and sulfate versus pH in leachates from mixtures of FBC residues and CSS 20 12 Molar concentrations of calcium and sulfate versus pH in extracts from mixtures of FBC residues and CSS 20 13 Concentrations of selenium in leachates from mixtures of FBC residues and CSS 23 14 Scanning electron micrographs of particles of bottom ash in an FBC residue 26 15 Scanning electron micrographs show the morphology of spherical and angular particles of fly ash from an FBC residue 27 16 Scanning electron micrographs of bottom ash particles from FBC residues 28 TABLES 1 Operating characteristics of the FBC plants 4 2 Percentage of FBC residue or aglime mixed with balance of CSS in wet-dry leaching experiments 5 3 Percentage of FBC residue or aglime mixed with balance of CSS in batch extraction experiments 6 4 Chemical composition of the FBC residues and coal slurry solids 8 5 Physical, chemical, and mineralogical properties of unleached FBC, aglime, and CSS samples 9 6 Peak areas from x-ray diffraction spectrometry for the major minerals relative to quartz at 20.85° 26, before and after wet-dry leaching 12 7 Chemical composition of solids after the 180-day batch extraction experiments 12 8 Chemical composition of solids after the wet-dry experiments 13 9 Change in the calcium carbonate content of the FBC residue, aglime, and CSS resulting from leaching or extraction 13 10 State of Illinois general use water quality standards and ranges of concentrations of various inorganic chemicals in leachates and extracts from mixtures of CSS and FBC residues 21 Cover Photo Discharge point near the containing berm of a coal slurry impound- ment. The sediment is predominantly coaly particles, but it also contains clay, pyrite, and other minerals. Editorial Board Jonathan H. Goodwin, Chair Michael L. Barnhardt Donald G. Mikulic Heinz H. Damberger William R. Roy Beverly L. Herzog C. Pius Weibel David R. Larson ABSTRACT The oxidation of pyrite in coal slurry solid, or tailings, from a coal preparation plant produces acidic leachate in the coal slurry impoundment. If left untreated, the acidic leachate may result in local environmental deterioration. The acidic solution could enhance the solubility and mobility of potential groundwater contaminants. If an alkaline material is added to the coal slurry solid, it either prevents or slows pyrite oxidation, or neutralizes the acid produced during the oxidation, and generally decreases the solubility of heavy metals. Therefore, the likelihood of groundwater contamination is diminished. Such codisposal of coal slurry solids with an alkaline material might allow revegetation of coal slurry solids without the soil cover presently required by regulatory agencies. In this research we investigated the interaction of water with mixtures of fluidized-bed combustion (FBC) residues or agricultural limestone (aglime) and coal slurry solid (CSS) under laboratory conditions. We compared the concentrations of various trace elements in leachates and extracts from the unmixed FBC and CSS materials and from mixtures of FBC residues and CSS materials with the Illinois standards for concentrations of trace elements in general-use water. We also used scanning electron microscopy to study particles of FBC fly and bottom ash. The major minerals in the unleached CSS were clay minerals and quartz (SiC>2). Also present were minor amounts of pyrite (FeS2) and calcite (CaCOa). The major minerals in the unleached FBC residues were lime (CaO) and anhydrite (CaSC^). When mixed with water, the lime was hydrated to portlandite [Ca(OH)2], which was converted to calcite during exposure to atmospheric carbon dioxide. The major ions in the leachates and extracts from the FBC residue-CSS mixtures were calcium (Ca 2+ ), sodium (Na + ), sulfate (S0 4 2_ ), and chloride (CI - ). The pH of the solutions from the FBC residues initially appeared to be controlled by portlandite, but later in the leaching or extraction period, the pH was probably controlled by the equilibrium between calcite, gypsum, and sulfuric acid. The alkaline species in the FBC residue effectively neutralized any acid produced by the oxidation of pyrite in the CSS. Constituents that were often observed in the leachates and extracts at concentrations greater than the respective concentrations for Illinois general use water quality were boron (B), chloride (CI - ), fluoride (F~), iron (Fe), mercury (Hg), manganese (Mn), nickel (Ni), selenium (Se), and sulfate (S04 2- ). Boron, chloride, cobalt, fluoride, manganese, and nickel were often present in leachates and extracts at concentrations that could be toxic to some plants. Scanning electron microscopy revealed that some particles of ash were enlarged during the FBC process by accretion of molten ash constituents, and that portions of some limestone particles were inhibited in their reaction with sulfur dioxide (SO2) in the FBC process, apparently due to the presence of magnesium (Mg). INTRODUCTION In the fluidized-bed combustion (FBC) of high-sulfur coal, pulverized limestone is injected simulta- neously with the coal to trap SO2 evolved during the high-temperature oxidation of pyrite in the coal. After combustion of the coal, the resulting residue typically contains substantial amounts of anhydrite (CaS04), lime (CaO) (Terman et al. 1978), and lesser amounts of coal ash. When mixed with water, the CaO in fresh FBC residue readily reacts with moisture to produce portlandite [Ca(OH) 2 ], which results in a strongly alkaline solution (pH 1 1 .5 to 13). Portlandite, on reaction with CO2 from the air, is converted to calcite (CaCOs). Previous researchers found that the alkaline nature of FBC residues makes them useful as substitute liming agents in agricultural practices (Terman et al. 1978, Bennett et al. 1978, Sidle et al. 1979, Korcak 1979, 1980, Stout et al. 1979a, b, Harness et al. 1987). Stout et al. (1982) concluded FBC residue appeared to be a satisfactory substitute for lime in the reclamation of acidic coal mine soils. These previous reports suggested FBC residue could be effective in buffering acid produced during the oxidation of pyrite in coal slurry solid (CSS). Calcite in water yields a solution with a pH of approximately 8. Several species of plants used in reclamation are tolerant to water of this pH. Coal slurry solid consists of coal, clay minerals, and heavy mineral matter, particularly pyrite, that are all rejected during coal cleaning. The CSS is discharged as an aqueous slurry to an impound- ment, where the solids settle out and the water is recycled through the cleaning process. If CSS and FBC residues can be disposed of together in a manner that allows plants to grow in the mixture, then coal mining companies could save money by avoiding the need to apply the soil cover presently required by state law in reclaiming coal slurry impoundments. Background research to characterize the materials was a necessary first step in the overall research project. The purpose of this study was to provide data on the (1) physical, chemical, and mineralogical characteristics of five FBC residues, an agricultural limestone (aglime), and two CSS materials; (2) composition of aqueous leachates and extracts from the as-received solids and mixtures of FBC residues or aglime and CSS; (3) major mineralogical changes that occurred during the leaching and extraction experiments; and (4) degree to which pyrite in the CSS would be oxidized in a system buffered by the addition of FBC residue. Because of time and financial constraints, replicate experiments were not conducted. This report is the result of two funded projects conducted during a 2-year period. During the first year, leaching experiments were conducted under cyclical wet and dry conditions in the laboratory to simulate codisposal of FBC residues and CSS in the unsaturated zone of the subsurface. A series of aqueous batch extractions was conducted during the second year to simulate the disposal of the materials in the saturated zone. The reported research is part of a larger program, the goal of which is to develop a method to reclaim coal slurry solid. In this reclamation method, FBC residues would be codisposed of with CSS. Two benefits would result from this codisposal: (1) acid generated during the oxidation of pyrite in the CSS would be neutralized by alkaline components of the FBC residue, and (2) the surface of the CSS-FBC deposit would be stabilized by plant growth. If this reclamation method is successful and is accepted by the prevailing regulatory authorities, it could alleviate the need for the thick soil cover. The overall research program is being accomplished in manageable steps, in the field and lab- oratory, for several years. Current federal law requires that a coal slurry impoundment be reclaimed after it is no longer used. One purpose of this reclamation is to prevent oxidation of pyrite in the CSS and eventual migration of acidic leachate to underlying groundwater. Illinois state law (Public Act 81-1015, "Surface Coal Mining Land Conservation and Reclamation Act") requires that coal slurry impoundments be revegetated after their use is suspended. If the coal slurry solid cannot support vegetation, it must be covered with a minimum of 4 feet (1.2 m) of soil capable of supporting vegetation. Pro- viding such a soil cover can be costly to the coal mine operator. At $1 .55 yd -3 ($2.03 rrr 3 ), it would cost approximately $10,000 per acre ($24,700 per hectare) to cover a slurry impoundment with 4 feet of soil. Other reclamation methods are available and may be permissible from a regulatory standpoint. Nawrot et al. (1985, 1989), for example, converted several coal slurry impoundments in Illinois into wetland habitats, which avoided both the use of soil covers and the production of acidic waters. Limestone was applied to the areas of the impoundment that contained the greatest amounts of pyrite prior to the planting of wetland plants. Our research focuses on the use of FBC residues rather than limestone in the reclamation of CSS, and the establishment of grasses and legumes normally associated with dry land environments. We believe this to be the first reported use of FBC residues in this manner. MATERIALS AND METHODS Grab samples of FBC residues (FBC-1-FBC-5) were collected from five Illinois FBC operators. All samples were collected from the outlets of storage silos or, in one case (FBC-4), from a freshly loaded truck. Three samples (FBC-1 , FBC-3, and FBC-4) were collected as composites of fly ash and bottom ash. The other two (FBC-2 and FBC-5) were collected as separate fly ash and bottom ash components. Portions of the two separate samples (FBC-2 and FBC-5) were combined prior to experimentation in a ratio of 25% bottom ash to 75% fly ash, according to the operators' estimates of ash production at the plants. The plants from which samples were collected had a relatively wide range of operating conditions (table 1). Each FBC residue was passed through a sieve with 3.36-mm openings (no. 6) prior to further sample preparation and characterization procedures. A grab sample of fresh CSS was collected from a relatively dry location (i.e., able to support the weight of a person and had no standing water) in an active coal slurry impoundment at two Illinois coal preparation plants. One sample (CSS-1) originated from the Illinois Herrin (No. 6) Coal and the other (CSS-2) originated from the Illinois Springfield (No. 5) Coal. A grab sample of aglime was collected at a limestone quarry in east-central Illinois. The sample was a slightly dolomitic limestone from the Silurian Racine Formation. The aglime was used as a control material with which to compare the FBC residues. The CSS materials and the aglime were air-dried by spreading them on polyethylene film and periodically mixing them by hand for about 2 weeks. All materials were split into smaller quantities by riffle sampling prior to their characterization and use. Characterization of the Unleached Solids The particle-size distribution of each unleached solid was determined by dry sieving each into seven particle-size fractions: <62, 62-125, 125-250, 250-500, 500-1,000, 1,000-2,000, and >2,000 urn. To use the aglime as a control for the FBC residues, we weighed portions of aglime in each of the particle-size ranges in the same proportions as determined for the corresponding FBC residue. The aglime portions were then combined and mixed, resulting in a material with approximately the same particle-size distribution as the corresponding FBC residue. The aglime portions were mixed so that particle size could be eliminated as a variable in comparing the behavior of the aglime with the behavior of the respective FBC residue. These aglime samples are referred to in this report as sized aglime. The bulk and particle densities of the CSS materials were determined by simple weight-volume relationships. The porosity of each CSS material was calculated from the bulk and particle densities. The pyrite content of the CSS samples was estimated by two methods: the peroxide oxidation method of Finkelman and Giffin (1986) and American Society for Testing and Materials Method D2492 for sulfur content in coals (ASTM 1990). The exchangeable acidity of the CSS samples was determined according to Method 9-4.2 of Thomas (1982). The calcium carbonate equivalent of the FBC residues was determined according to American Society for Testing and Materials Method C602 (ASTM 1 990). The CaO content of the FBC residues was estimated as follows. A weighed sample of the FBC residue was treated with deionized water until the sample was saturated. The sample was oven-dried at 110°C and mixed periodically to prevent cementation, then it was cooled and weighed. We assumed the weight gain to be equal to the amount of water absorbed in the hydration of CaO to Ca(OH)2. This assumption allowed for a back calculation of the amount of CaO in the fresh FBC residue. The assumption was believed to be valid because (1) no gypsum was observed by X-ray diffraction spectrometry (XRD) in two FBC residue samples after storage for 5 months; hence, anhydrite was not easily hydrated; and (2) the content of other potentially hydratable minerals, such as clay minerals, was low. An estimate of the proportion of Ca-containing species in each unleached FBC residue was calculated from the calcium carbonate equivalent, total Ca, sulfate-S, the carbonate-carbon content, and the amount of water consumed upon hydration. In making these estimates, we assumed that only CaO, Ca(OH)2, and CaC03 contributed to the observed calcium carbonate equivalent, all SO4 2 " occurred as CaS04, and all CO3 2 " occurred as CaC03. We also assumed that all lime was hydrated to portlandite during leaching or extraction, because no lime was detected by XRD in the leached or extracted residues. Each unleached and leached solid was analyzed by instrumental neutron activation anaiyt for arsenic (As), cobalt (Co), chromiun (Cr), manganese (Mn), and selenium (Se) (Harvey et al. 1983); by optical emission spectrometry (OEP) for silver (Ag), boron (B), beryllium (Be), ana vanadium (V ) (Harvey et al. 1983); and by atomic absorption spectrometry (AAS) for cadmium (Cd), copper (Cu), nickel (Ni), lead (Pb), and zinc (Zn) according to ASTM Method 3683. Major and minor elements in coal and coke ash were analyzed by wavelength-dispersive X-ray fluorescence spec- trometry (XRF) for aluminum (Al), barium (Ba), calcium (Ca), iron (Fe), potassium (K), magnesium (Mg), sodium (Na), sulfur (S), and silicon (Si). In the XRF method, a 125-mg aliquot of an ashed sample of solid was fused with 1 g of a 50:50 mixture of lithium metaborate and lithium tetraborate in a platinum crucible for 15 minutes at 1,000°C. The fused mixture was pulverized with 2% Somar m 6 CD LL ? o CO LL. CO 6 CD LL CM 6 CO LL w ♦* c CO Q. ^ o ■ o CD m LL LL CO XT •^ O 8 to 'C o 2 1 CO -C to CJ Q) c O CO {" CO CO Q. u O c T— CO c 0) CD n Q. O 1- c 3 CM, O) "O ^ C a> m CO CO o 3 c z O c i— Q. =' o CO 55. CO CO E 3 E E E E xl°> — . m -A CM co m o> in CO CO o 8 CO I in CD *- co a. o CO CO o CO >. (V O CO LL < CO in cn d c z •c a. — f CO c^ Q. E E 8 o CO 0) to o "* c in c r L CM o> CO X) c\i E *^ T— CO 2 A in o!5 h» 3 CM CO i §2 w d m i- CO CM 1"*. co- m * «q i- V CM* TT 6 CO •*' co co in co 'co cd in c o o.= .£ LL CO =< _l C »- c c Jm CO CO O O c o S> c ^ O 3 CO LL CO 2 ^ .9 a) m JS o> d 3 C Z y c . •-= Q.= O CO =. c 3 in co c E E o E h-; r» O) CM CM in 1 T— 6 ■* VI V CM x: x: cn — cp 5^ « •CM 3 O t- CO CO CM — . CO CO c 1 %| I => -C Z CD P *- c ..= CO — C O X =. J a> Q. co °> ._ _ o c c CO I «> to T *' ~>CM I! o a5 CO Q. CT> i_ y- XT o o a> CO ♦— k* o. 2 3 "c0 2 CO CO CD o E CL E 3 co w CO co gi 3 CD L) '5 3 (0 a> c g a. c fl) c 1 (0 c CO 3 c Q. CO 1 u CO 1 3 O •a CO CO 0) .a X) E "CO 0) >, o o (0 o o CO CO O o o CL h- O CO a. o co O LL O CO O O 3 Q. C co CO o> O ro to c O o CO LL CO N CO o CD LL O 0) E 01 i_ 3 t) ,C0 *3 C co 2 fc 2 ® - cu o Q. O CO Q. O 3 O o.f °x: co .h x: co CO < < CO CO c o o o CO CO CO CO .CO •- co o- co £ QJ N D. d) ro 0) 3 E £ § 2 s O) o CO O 3 CO LU CO £ Q) (O CO CO 8 E x: o a> c CO O CO CD o>|cl| 3 3 x: CO CM 1^ CO o o o o CO CO in in CD x: x: CO CO "CO a> cb o» E o o in CO a. 40,0 ry:2 & to o u. CO CD Jg N C •• CO in ^c (0 CO F CO E X3 O u. 8^ m 8 x: E S § E x: o CD c C CD f*- co- CD (0 CO O o o Q. CO CO in a < CO CO lu cn c CO CO CO E CD CO CM x: E CO o cb xf CO j; >N O 5= X3 o o ^— CO .. CO •S E CO o co in o CO >« o o CO -t CD LL JD in s - in CM 52 -c ♦* CO CO -= *t CM d CO o co x: o y- CO .. CO o "S E o CO CO o -^ o CO CO CD LL XJ d 00 c O CL c g o o o x: co CO 0) X) co > < E o E CO x: CO CO ^» CO to .2 JZ CD E ^ ^_ O) CO ^ CO R -o CD CO T) s ^ CO o co o ^_, ■«— ' c c CD 3 k. o 3 E o < T3 CD r a CO ha o c mix in a tungsten carbide grinding vial on a No. 6 Wig-L-Bug® 1 . The ground sample was pressed into a disk, 1 inch in diameter, at 40,000 psi. The pressed disk was then analyzed quantitatively on a Rigaku Model 3371 X-ray fluorescence spectrometer. The carbonate-carbon content of each sample was determined by using a UIC® carbon analyzer (Cahill and Autrey 1988). The solids were analyzed for bulk mineralogical composition by XRD (Hughes et al. 1994). In addition, samples of unleached FBC fly ash and bottom ash were examined by scanning electron microscopy with energy-dispersive X-ray spectrometric (SEM-EDS) detection. Preparation of Mixtures The amount of each FBC residue added to each CSS material was calculated according to the acid-neutralizing ability of the FBC residue and the net acid-generating potential of the CSS material. (For the wet-dry leaching experiments, the net acid-generating potential of the CSS material was derived from the Finkelman and Giffin [1986] method, which probably underestimated the amount of FBC residue required to stoichiometrically match the amount of pyritic sulfur in the CSS. Because calcite was not fully consumed in the wet-dry leaching experiments, the underestimation did not create a serious deficiency). The calculation was based on the reaction between CaC03 and the acid produced during pyrite oxidation. Four moles of CaC03 are consumed for each mole of FeS2 oxidized, according to the following reaction: FeS 2 + 3.750 2 + 3.5H 2 + 4CaC0 3 = Fe(OH) 3 + 2CaS0 4 + 2Ca(HC0 3 )2 (1) Wet-Dry Leaching Experiments Laboratory leaching experiments were prepared by mixing FBC residue or sized aglime and CSS in the proportions calculated as described (table 2). Each of 20 mixtures was placed into a plastic flower pot, 17.8 cm in diameter. The bottoms of the pots were lined with filter paper to prevent the elution of fines. The eight unmixed samples (five FBC residues, one aglime, and two CSS materials) were also leached. Each flower pot was nested into a polypropylene beaker. We added approxi- mately one pore volume (200-500 mL) of deionized water weekly to each mixture for 1 5 weeks. This interval allowed for at least partial drying of the solids and oxidation of pyrite, yet allowed the research to proceed at a reasonable pace. The water was allowed to flow through the mixture by gravity, and leachate was collected in the plastic beaker. The first pore volume of water produced no leachate because this amount was re- quired to initially saturate the solid. Leach- ates from each two pore volumes of water were composited to accumulate sufficient volume for the desired analyses. No attempt was made to prevent drying of the solids between additions of water. Batch Extraction Experiments Each of three FBC residues (FBC-3, FBC-4, and FBC-5) or sized aglime was mixed with each CSS in the percentages shown in table 3. Only FBC-3, FBC-4, and FBC-5 were used in the batch extraction experiments because FBC-1 and FBC-2 appeared to be similar to FBC-5. Extractions from these mixtures were conducted for 3, 10, 30, 90, and 1 80 days. One part by weight of solid was mixed with four parts by weight of deionized water. The mixtures were agitated briefly and aerated by flowing air into the solution through a frit Table 2 Percentage of FBC residue or aglime mixed with balance of CSS in wet-dry leaching experiments. FBC residue or aglime mixed with CSS-1 FBC-1* FBC-2 FBC-3 FBC-4 FBC-5 2.1 2.8 1.7 1.4 2.7 FBC residue or aglime mixed with CSS-2 FBC-1 FBC-2 FBC-3 FBC-4 FBC-5 7.6 10.0 6.4 5.3 9.9 *Or aglime with size composition adjusted to approximate that of the corresponding FBC residue The use of trade names in this report is for descriptive purposes only and does not constitute endorsement by the Illinois State Geological Survey. for 4 hours per workday to encourage pyrite Table 3 Percentage of FBC residue or aglime mixed oxidation. Water that evaporated during the with balance of CSS in batch extraction experiments, experiments was replenished weekly. The pH of each solution was determined prior to filtra- FBC residue or ag , jme mixed wjth css _., tion at the end of each extraction period. Each FBC residue or aglime mixed with CSS-2 filtrate was analyzed chemically; each ex- FBC-3 0.35 tracted solid was air-dried and analyzed FBC-4 0.29 chemically and mineralogically. 1 * ' FBC-5 0.54 Analysis of Leachates aglime 0.29 and Extracts — The leachates and extracts were analyzed by — inductively coupled plasma emission spec- FBC-3 32.67 trometry (ICP) for aluminum (Al), boron (B), FBC-4 27 33 barium (Ba), calcium (Ca), cadmium (Cd), chromium (Cr), copper (Cu), iron (Fe), potas- FBC-5 50.64 sium (K), magnesium (Mg), manganese (Mn), aglime 26.81 sodium (Na), nickel (Ni), lead (Pb), vanadium (V), zinc (Zn), and zirconium (Zr). Inductively coupled plasma-mass spectrometry (ICP-MS) was used to determine As and Se in the wet-dry leachates. Graphite furnace atomic absorption spectrometry (GF-AAS) was used to determine As and Se in the batch extracts. Both sets of analyses for As and Se were conducted by the Illinois Hazardous Waste Research and Information Center (HWRIC). Chloride, fluoride, and sulfate were determined by ion chromatography (IC), and alkalinity was detemined by titration using standardized dilute hydrochloric acid. The Eh and pH were determined by conventional electrode methods. The platinum electrode-reference electrode pair used for Eh measurement was calibrated against a standard ZoBell solution (ZoBell 1946). RESULTS AND DISCUSSION Physical, Chemical, and Mineralogical Characteristics of FBC Residues and CSS Particle size The particle-size distributions of FBC-1 , FBC-2, and FBC-5 were similar (fig. 1 ); 50% or more of the particles in these residues was smaller than 62 urn. The particle-size distribution of FBC-3 was nearly uniform; about 50% of the particles was smaller than 250 urn. More than half the particles in FBC-4 was smaller than 500 urn. The particle-size distribution of the aglime was similar to that of FBC-4. Approximately 50% of the particles in CSS-1 was smaller than 125 urn, and about 50% of the particles in CSS-2 was smaller than 500 urn. It is expected that the smaller the average particle size of an FBC residue, the more quickly it will dissolve on exposure to moisture. In codisposal, small particle sizes will allow the FBC residue to respond quickly to an acidic leachate, but it might then be necessary to replenish the FBC residue every few years. Chemical composition The chemical composition of each of the five FBC residues in this study appeared to be comparable with those of previous studies (table 4). The ranges of constituent concentrations reported by others for FBC residues are listed for comparison (Bennett et al. 1978, Sidle et al. 1979, Harness et al. 1987, Fennelly et al. 1987, and Tavoulareas et al. 1987). The concentrations reported by others for elements in Illinois coal ash (Gluskoter et al. 1977) and in limestones (in general) (Brownlow 1979) are also listed for comparison. The concentrations of several elements, for example boron, in the FBC residues were higher than the concentrations of those elements reported for limestone. Coal ash, with its typically higher concentrations of trace elements, probably contributed to the increased concentrations observed in the FBC residues. Mineralogical composition The dominant crystalline phases in most of the FBC samples were portlandite [Ca(OH)2], anhydrite (CaSC^), lime (CaO), and to a lesser extent quartz (Si02; table 5). In the case of FBC-3, the fluidized-bed combustion boiler was used to supply heat to a complex of buildings from October 1991 through April 1992. After April, the ash silo was emptied, but a small amount of ash clung to the walls of the silo. A portion of this remainder was collected in September 1992. Hence, it had been exposed to the atmosphere for approximately 4 to 5 months. As a result, FBC-3 consisted largely of CaC0 3 and Ca(OH) 2 (table 5). o 100 <62 62 125 250 500 1000 2000>2000 <62 62 125 250 500 1 000 2000 >2000 Sieve opening (pm) c O 70 k_ 0) a. <- 60 O) > "to 40 Zl E 3 30 o 20 10 100 90 80 +4 c o 70 L— o Q. I: 60 O (V $ 50 a> > ro 40 3 F 3 30 o 20 10 <62 62 125 250 500 1000 2000>2000 0^ J I I L _L -e- css-1 -B- CSS-2 J I l_ <62 62 125 250 500 1000 2000 > 2000 Sieve opening (pm) Figure 1 Particle-size distribution of FBC residues, aglime, and CSS materials. On the basis of the calculated results, the FBC residues (except for FBC-3) consisted largely of CaSCM and unreacted CaO (table 5). On exposure to atmospheric moisture, CaO is easily hydrated to Ca(OH)2 (Sisler et al. 1 961 , Hurlbut 1 961 ). Evidence of such hydration during the 5-month storage of a sample in the laboratory is illustrated in figure 2 by the decrease in the peak area of lime and the increase in the peak area for portlandite. Portlandite readily reacts with CO2 to form CaC03 (Sisler et al. 1961). The FBC residues also contained coal ash, which typically contains large amounts of Si02 and AI2O3. Whereas others (Graham and Hower 1994, Bland et al. 1995) have observed the presence of ettringite [Ca6Al2(S04)3(OH)i2 # 26H20] in FBC residues, we detected it only in FBC-3. The CSS materials contained varying amounts of pyrite and ash-forming minerals. By inference from chemical analysis, the aglime consisted principally of calcite (CaC03) and dolomite [CaMg(C03)2], but it also contained quartz (Si02; <7.0%), hematite (Fe203; iO.62%), and minor amounts of clay minerals. Both CSS materials contained calcite (table 5) and, therefore, had © os c 2 « O CO OS CO __ CO .}= CO fe c o > 8 ° 8.S s™ S 3S o to to to TJ © c © TJ to c o c O c o O CM CO CO o (0 CO o CD E O) m i o co o m CO 6 co C\i 6 CO O CO u. to © 'S Q. CO * * § 7 7 7 n_ i s i c CO CM k 5! to ^ l _dd7»->->-'- C I t- c c c c CO CM O i- JZ I = CD CO ^ Si S CO * < t- CO CM CM If) m o _? ^ *f\ ^\ r^ <-"\ w 00 CM CM CO O «= c c J 6 c s o in i o CO CO o CO W os 7 CM CM T CO CO CO O tJ- £ 7 7 T 7 q CO CO OS CO CM ^ CO CO o CO 2 co o oo i- 00 ? > CO 7 ■ CM OS n oo N J! ci CO m OS i^r^fS'f-CMcomr^m'*^- OT-^OOT-Ot-T-T-COT-'*'^ in t- co in co N S * ? ??ss i? 8 8 ? fc 8 8 S U 5 8 8 58 5 J M5 8 r; B n ? s « • q S2incodcodd§«^d-cod^^w-- v c5 ^ v cm co * ® § s si a 1 § g | || h. T- o ^ TJ c TJ C TJ C CM CM ZL o ZZ t- y- v ^ ° ,- co m v co v co o I s - "> _► — os ^ £ d 2 2 v v V l"» ^ ui fi 00 3; i- 1^ CM •5 ^ O 00 CO CM ■* US CM 00 CO O t- T- *"" CO CO 00 ,»> .*» • ▼- z: Jv co in IT n »- m w v m in 5 r^ co to t-: oo CM g 55 co to ^ oq g ^ J d d d 5 d co co cm co co d 8 £ CO r^ S «» CO CO m o m CO i- o in oo - O) O N O) N t Jl 00 V CM i- y- t- CO 00 CM 3 g o s 00 y- O) 00 OS t- CO CM CO d 81 - CM ■* w CM 5° ^ S cOT-oocoinco.rtQo,- _ o 5 ™ 5 s? a CO CO in pj o to co co ^ *" v v y- °* y- ^uiOti^dciooiOiod CO o co r^ N 9 CD £ oo CO in co § § CO N 00 CO CO CM y- t- CO CO w m os os in co o CM co oo CO CO 3" y- o 4 Tf CO oo co cm m 00 * CO CM OS m ™ » n t n to ^mini^JGdddosdddddcMco'^cMco v-^coi-co^ ? 8 C\J ■^ CM in yf O) O) 5> Q> CJ) O) O) CO < o a> 9, ° O) CO 2 O 2 O to >. < 0- O) O ir O to ^ O h- < OS JC X^W W co £0 ~OoEiI > Ef > E£|P g mcQOOO22C0>N oo O) CD to CO CD i_ JO 3 o > CO H- C CO 00 OS CO CD c c 0) u. _ - CO co °2 o> ZL. co c5 Q) >n to S CO C 0) CO E I CO -n I c . CO IP o> s; ® c5 ^ © i_ o> CO o^ x» o co2 c &°i ^T O CO co^ E © | 9 "Is E id o 2 co > © o c to J E O T3 O £ S o — CO ■6 £ o to E *5 Q) © © t E E o © = CO d) c o to .9 co O) CO O < CC TJ n (!) T" u O 0) IX CD CD ■n O C o c ^ TJ C C Table 5 Physical, chemical, and mineralogical properties of unleached FBC, aglime, and CSS samples. FBC-1 FBC-2 FBC-3 FBC-4 FBC-5 Aglime CSS-1 CSS-2 Physical and chemical properties Bulk density (g cm" 3 ) » — — — — — 0.84 1.51 Particle density (g cm" 3 ) — — — — — — 1.35 2.46 Porosity — — — — — — 0.38 0.39 Pyrite (%) f — — — — — — 3.66 12.4 Pyrite (%)* — — — — — — 3.01 24.7 Exchangeable acidity (meq kg" 1 ) — — — — — — 16 23 CCE (%) 65.1 48.2 77.8 93.0 50.2 94.8 — — CaC0 3 (%) 3.58 4.00 39.8 9.75 3.67 93.1 4.83 15.9 Calculated proportions of calcium compounds (%) i CaO 28.6 23.4 <0.1 45.6 22.7 — — — Ca(OH) 2 7.73 1.74 28.1 1.36 4.41 — — — CaS0 4 § 27.4 25.7 12.9 26.6 24.7 — — — CaMg(C0 3 ) 2 — — — — — 19.2 — — Minerals detected by XRD Portlandite M M M M M nd" nd nd Quartz M M t m M m M M Anhydrite M M M M M nd nd nd Calcite m m M m m M m M Lime M M nd M M nd nd nd Ettringite nd nd t nd nd nd nd nd Hematite m M m m m nd nd nd lllite/smectite nd nd nd nd nd nd m m lllite nd nd nd nd nd m m m Kaolinite nd nd nd nd nd nd m m Gibbsite nd nd nd nd nd nd t t Plagioclase feldspar nd nd nd nd nd nd m m Pyrite nd nd nd nd nd nd m M Dolomite nd nd nd nd nd M nd nd f Method of Finkelman and Giffin (1986) *ASTM Method D2492 § The amount of CaS0 4 in each FBC residue was calculated from the value for Sul-S (table 'M = major component (10-100%), m = minor component (0.5-10%), t = trace component * — = not determined (not sought) ** nd = not detected (sought but not observed) 4) (<0.5%) some natural buffering capacity toward acid. On the basis of the number of chemical equivalents of calcite and pyrite present, and considering that the oxidation of 1 mole of pyrite will consume 4 moles of calcite (Eq. 1), there was enough calcite in CSS-1 to neutralize one-half the acid-generating potential of the pyrite present, and enough calcite in CSS-2 to neutralize one-fifth of the acid-gener- ating potential of the pyrite present. The major changes in the mineralogy of the FBC residues during either the leaching experiments or the batch extractions, as revealed by X-ray diffraction spectrometry, were the general decrease in lime (CaO) and portlandite [Ca(OH)2)] content and the increase in calcite (CaC03) content. The relative height of the primary peak in the X-ray diffractogram for portlandite increased on storage of the FBC residues. Figure 2b shows the X-ray diffraction spectra for unleached FBC-4 at 4 months (January 31, 1992) and 7 months (July 2, 1992) after collection, and for leached FBC-4. The peak height (relative to quartz at 20.85° 26) for portlandite at approximately 18° and 34° 20 became pro- gressively greater through time, whereas the peaks for lime collapsed. Note also that the portlandite peaks in the spectrum from July 2, 1 992, were rather broad, an indication that the portlandite crystals CPS a 3888.8- 2788.8- 2488.8- 2188.8- " * ■»' w*mww ^ ^-^»»*« w 1888.8i c o ~ 1588.8- > "i 1288.8- 988.8- 688 . 8- 388. 1 FBC-1 (7/2/92) £ U^iAilJU 4e Degrees 26 CPS b 2888.8- 1888.8- 1688.8" 1 488 . 8- w 1288.8-1 Q) ■J 1888.8- > -55 888 . 8- CE 688. 0H Co - .B FBC-4 ( leached) . FBC-4 (7/2/92) J^ li 4e Degrees 20 Figure 2 X-ray diffraction spectra of FBC-1 and FBC-4. (a) X-ray diffraction spectra show the decrease in lime (Ca) content and the increase in portlandite (P) content of FBC-1 after storage for 5 months, (b) X-ray diffraction spectra show the decrease in lime content and the increase in portlandite content of FBC-4 after storage for 5 months, and the refinement of the crystallinity of portlandite, the disappearance of anhydrite (A), and the appearance of bassanite (B) during leaching. Calcite (Cc), iron oxides/hydroxides (H), and quartz (Q). 10 in the specimen were small and/or disordered. By comparison, the portlandite peaks in the spectrum for the leached FBC-4 were narrower and showed increased relative peak height, which is indicative of larger crystal size as a result of the wet-dry leaching. The calcium carbonate equivalent of a lime-based material is an indicator of its acid-neutralizing ability. The calcium carbonate equivalent values of the FBC residues indicated they had relatively large acid-neutralizing capabilities (table 5). The acid-neutralizing components expected to be present in the fresh FBC residues were CaO, Ca(OH)2, and MgO. Most of the CaC03 in the feed limestone was calcined during fluidized-bed combustion. If MgCCh and CaMg(CC>3)2 were present in the feed limestone, they too would have been calcined. The calcium carbonate content in all the as-received FBC residues, except FBC-3, was less than 10% (table 5), indicating little exposure to atmospheric carbon dioxide had been allowed prior to sample collection. Effects of Leaching The major changes in the mineralogy of the FBC residues during either the leaching experiments or the batch extractions, as revealed by XRD, were the general decrease of lime and portlandite content and the increase of calcite content. The relative peak areas for various minerals in the FBC residues and aglime before and after wet-dry leaching are given in table 6. In each of the FBC residues in which lime was initially present, none was detected after leaching. For FBC-2, FBC-3, and FBC-5, the relative peak height for portlandite was substantially decreased after leaching, and in every case, the peak area for calcite was significantly greater after leaching than before. As discussed earlier, the increase in the relative peak height for portlandite during the leaching of FBC-1 and FBC-4 was apparently the result of improved crystallinity, rather than increased portlandite content. On the basis of the data in table 6 for lime, portlandite, and calcite, and in tables 4 and 8 for inorganic carbon, the lime or portlandite in the FBC residues appeared to have reacted with carbon dioxide from the air to form calcite. Mineralogical changes in the residues from the batch extractions were minor and nonsystematic. The chemical composition of the FBC, CSS, and aglime samples after 180 days of extraction (table 7) was similar to the chemical composition of the wet-dry leached samples (table 8). In most of the leaching experiments (table 9), the calcium carbonate content increased, whereas in the batch extractions it decreased. Iron oxides and hydroxides were present in all the FBC residues as a result of the incorporation of coal ash during the combustion cycle. Chemical analyses indicated the pyrite content of CSS-1 was approximately 3.01%, whereas that of CSS-2 was 24.7% (table 5). The greater density of pyrite (5.00 g cm -3 ) compared with that of coal (approximately 1.3 g cm -3 ) and the greater content of pyrite in CSS-2 than in CSS-1 were responsible for the higher bulk and particle densities of CSS-2. The fractional porosities of the materials, as packed in the leaching containers, were approximately equal, 0.38 for CSS-1 and 0.39 for CSS-2. In the presence of certain bacteria, pyrite can be oxidized rapidly under acidic conditions. A goal of this research was to determine whether pyrite would be oxidized under the buffered conditions presented by the admixture of FBC residue to CSS. We observed the pyritic sulfur content decreased from 1.61% to 1.43% in CSS-1 and from 13.2% to 11.5% in CSS-2 during the wet-dry leaching experiments. These percentages amount to relative decreases in pyritic sulfur content of 11.2% and 12.9%, respectively. In the mixtures of CSS and FBC residue or aglime, the relative decrease in pyritic sulfur during the wet-dry leaching experiments ranged from 7.8% to 1 9.6%. These data gave a clear indication that pyrite in the solids was oxidized during the wet-dry leaching experiments under the pH-buffered conditions provided by the FBC residue. Gypsum (CaS04 # 2H20), the product of reaction between calcite and sulfuric acid, was observed by XRD analysis of CSS-2 (fig. 3). Anhydrite (CaS04), the high-temperature reaction product of SO2, CaO, and O2, was observed in the unleached FBC residues (fig. 3), but gypsum was not. We expected gypsum would form in mixtures of CSS-2 and FBC residues during leaching, from the reaction of the alkaline, calcium-containing constituents of FBC residues with the sulfuric acid generated during the oxidation of pyrite in the CSS. After leaching, however, neither gypsum nor anhydrite was detected in the mixtures. Rather, bassanite (CaS04 ,1 /2H 2 0) was detected in all mixtures that contained CSS-2 (figs. 2b, 3). It is unlikely that anhydrite was partially hydrated to bassanite during the leaching experiments because this reaction proceeds very slowly, if at all (Hardie 1967). The hydration of anhydrite to gypsum is also slow. Therefore, we attributed the disappearance of anhydrite from the mixtures to dissolution during the leaching experiments. 11 Table 6 Peak areas from x-ray diffraction spectrometry for the major minerals relative to quartz at 20.85°28, before and after wet-dry leaching. Sample Portlandite Anhydrite Calcite Lime Ettringite Hematite Bassanite Before leaching FBC-1 0.23 1.50 nd* 2.17 nd 0.23 nd FBC-2 0.13 0.60 0.06 0.76 nd 0.40 nd FBC-3 1.93 1.53 5.51 nd 0.20 0.26 nd FBC-4 0.52 2.58 0.98 9.80 nd 0.36 nd FBC-5 0.13 0.85 nd 0.99 nd 0.19 nd Aglime nd nd 12.12 nd nd nd nd After leaching FBC-1 0.65 0.06 1.32 nd nd 0.26 0.23 FBC-2 nd 0.07 0.92 nd nd 0.53 0.24 FBC-3 0.57 0.08 6.03 nd nd 0.30 0.19 FBC-4 4.74 1.80 5.39 nd nd 0.49 0.69 FBC-5 0.03 0.03 0.81 nd nd 0.21 0.15 Aglime nd nd 10.35 nd nd 0.51 nd *nd = not detected Table 7 Chemical composition of solids after the 1 80-day batch extraction experiments. Concentrations determined in this study Species FBC-3 FBC-4 FBC-5 Aglime CSS-1 CSS-2 Si0 2 (%) 7.96 5.61 18.91 7.02 20.06 29.10 Al 2 3 (%) 1.81 1.54 4.76 1.30 5.84 7.29 Fe 2 3 (%) 2.96 2.38 5.56 0.64 2.99 21.34 MgO (%) 0.58 0.87 0.91 4.28 0.40 0.76 CaO (%) 43.36 49.19 30.16 45.42 2.73 9.42 Na 2 (%) 0.06 0.06 0.19 0.04 0.12 0.34 K 2 (%) 0.16 0.10 0.34 0.52 0.88 1.27 MnO (%) 0.17 0.84 0.65 0.35 0.26 0.99 As (mg kg" 1 ) 4.5 7.3 10 2.0 2.7 85 B (mg kg" 1 ) 85 114 300 <10 113 71 Ba (mg kg' 1 ) '40 <10 230 30 145 315 Cd (mg kg- 1 ) <3 <4 <3 <3 <1 28 Cr (mg kg" 1 ) 22 17 47 5.8 36 67 Cu (mg kg" 1 ) 16 20 25 11 17 38 Mo (mg kg -1 ) <10 <10 14 <10 5 18 Ni (mg kg" 1 ) 21 25 34 17 28 72 Se (mg kg' 1 ) 1.4 2.8 5.1 <0.5 2.8 14 V (mg kg" 1 ) 41 54 125 16 33 54 Zn (mg kg" 1 ) 122 79 110 <17 52 1276 Ash (%) 64.71 74.05 74.90 59.61 36.37 80.46 Pyr-S (%) * — — — 1.37 12.52 Sul-S (%) 2.59 3.88 4.66 — 0.21 1.35 Org-S (%) — — — — 1.96 1.86 Tot-S (%) 2.84 5.10 4.99 0.03 3.54 15.73 In-C (%) 7.17 3.35 2.48 11.15 0.46 1.76 * — = not determined 12 Table 8 Chemical composition of solids after the wet-dry leaching experiments (1 5 pore volumes). Concentrations determined in this study Species FBC-1 FBC-2 FBC-3 FBC-4 FBC-5 Aglime CSS-1 CSS-2 Si0 2 (%) 15.92 19.07 8.30 5.27 19.08 6.93 20.08 29.58 Al 2 3 (%) 5.16 5.40 1.87 1.37 4.89 1.20 5.98 7.54 Fe 2 3 (°/o) 4.77 10.22 3.20 2.27 5.81 0.61 2.98 20.58 MgO (%) 0.91 0.76 0.57 0.83 0.88 4.04 0.36 0.68 CaO (%) 28.66 30.58 44.82 50.76 30.62 45.39 2.42 9.02 Na 2 (%) 0.36 0.15 0.16 0.07 0.07 0.16 0.10 0.36 K 2 (%) 0.35 0.35 0.16 0.11 0.35 0.53 0.89 1.30 MnO (%) 0.08 0.06 0.02 0.09 0.07 0.04 0.03 0.09 As (mg kg" 1 ) 8.9 9.3 5.1 7.3 10 2.0 2.7 78 B (mg kg' 1 ) 312 373 153 191 433 <10 135 48 Ba (mg kg" 1 ) 85 108 40 49 81 10 39 58 Cd (mg kg" 1 ) <2.2 <2.2 <2.2 <2.4 <2.0 <2.3 <0.9 15.6 Cr (mg kg" 1 ) 36 63 24 17 49 6.7 37 69 Cu (mg kg" 1 ) 23 17 6 11 21 <7 15 37 Mo (mg kg" 1 ) nd* 9 nd nd 7 nd nd 29 Ni (mg kg" 1 ) 23 43 <19 <20 38 <20 20 56 Se (mg kg" 1 ) 3.9 8.7 1.3 2.2 5.3 <0.5 2.1 11 Vfmgkg" 1 ) 85 108 40 49 81 10 39 58 Zn (mg kg" 1 ) 405 182 97 86 125 32 74 1060 Ash (%) 81.35 81.33 67.60 74.83 76.11 59.82 36.43 80.66 Pyr-S (%) 0.01 0.01 0.03 0.01 0.02 <0.01 1.43 11.54 Sul-S (%) 5.55 5.20 2.62 4.90 5.12 <0.01 0.07 2.36 Org-S (%) 0.41 0.55 0.02 0.51 0.47 <0.01 1.73 0.75 Tot-S (%) 5.97 5.76 2.67 5.42 5.61 <0.01 3.23 14.65 In-C (%) 2.85 2.71 6.21 3.58 2.39 11.03 0.81 1.49 Org-C (%) 2.46 4.86 5.46 3.66 10.05 0.23 47 7.58 Tot-C (%) 5.31 7.57 11.67 7.24 12.44 11.26 48 9.07 * nd = not detected Table 9 Change in the calcium carbonate content of the FBC residue, aglime, and CSS resulting from leaching or extraction. Change in CaC0 3 content (%) Sample Wet-dry leaching Batch extractions (180 days) Unmixed FBC residue +12.0 to +20.2 +10.19 to +11.94 Aglime -2.90 -0.95 CSS-1 +1.40 -0.95 CSS-2 -3.80 -0.10 Mixtures CSS-1 + FBC residue +0.23 to +2.47 -0.65 to -0.70 CSS-1 + aglime 0.00 to -1.40 -0.85 CSS-2 + FBC residue +1.00 to +2.60 +3.10 to +6.40 CSS-2 + aglime -2.00 to -3.60 -0.49 13 CPS 4806 Degrees 26 Figure 3 X-ray diffraction spectra of FBC-1 and CSS-2 before wet-dry leaching and CSS-2/FBC-1 after 15 pore volumes of water were leached through the mixture. Anhydrite (A), bassanite (B), calcite (Cc), clay (020), gypsum (G), illite (I), iron oxides/hydroxides (H), kaolinite (K), lime (Ca), marcasite (Ma), plagioclase feldspar (Pf), portlandite (P), pyrite (Py), and quartz (Q). Gypsum probably precipitated during the leaching experiments, but it was partially dehydrated to bassanite during low temperature drying of the samples at 70°C in preparation for XRD analysis. Deer et al. (1962) indicated that gypsum is dehydrated to bassanite, beginning at a temperature of about 70°C. Bassanite was observed only in mixtures that contained CSS-2, probably because the bassanite content was above the detection limit only in those mixtures. The amounts of FBC residue added to CSS-1 were 1.4% to 2.8% and the amounts added to CSS-2 were 5.3% to 10.0%. The greater the amounts of pyrite and FBC residue in the mixture, the greater the probability of gypsum formation during the wet-dry leaching, and hence the greater the amount of bassanite on drying of the sample. pH of Leachates and Extracts Lindsay (1979) indicated that in a soil containing both gypsum and calcite at equilibrium under a fixed partial pressure of CO2, the addition of sulfuric acid will cause calcite to dissolve and gypsum to precipitate, according to the following reaction: H2SO4 + CaC0 3 + H 2 =» CaS0 4 «2H 2 + C0 2 f (2) As long as there is a supply of CaC03, the composition of the solution phase will not change with the addition of sulfuric acid, and the pH of the solution will be approximately 7.8 at a CO2 partial pressure of 0.0003 atmosphere, the partial pressure of CO2 in the atmosphere (Lindsay 1979). This stoichiometric equation above represents the probable reaction in the neutralization by calcium carbonate of sulfuric acid generated during pyrite oxidation. The pH of the wet-dry leachates from the unmixed FBC residues ranged from 10.5 to 13.0 (fig. 4a), and that of the batch extracts ranged from 11.9 to 12.6 (fig. 4b). The range of pH values in the solutions from the unmixed aglime was 7.7 to 8.4 in the wet-dry leachates and 7.5 to 8.4 in the batch extracts (fig. 4). The high pH of the solutions from the unmixed FBC residues was characteristic of portlandite dissolution, which would produce a solution with a pH of approximately 12.4. The pH of the solutions from the unmixed aglime was characteristic of a solution in equilibrium with calcite. 14 14 12 - 10 - 1 8 i i i r "/ A A ^V (a) 10 12 14 16 Pore volumes leached 60 90 120 Extraction time (days) 180 Figure 4 pH in leachates and extracts from unmixed FBC residues and CSS. Unmixed CSS-1 initially contained 4.83% CaC03, which buffered the pH of leachates to about 8 (fig. 4a) throughout the wet-dry leaching period. The pH of 8 indicates the CSS-1 -water system was probably buffered by calcite. Leachates from unmixed CSS-2 (fig. 4a) were acidic at the beginning and the end of the leaching period. This material contained 15.9% calcite at the beginning of the experiments. The reason for the decrease in pH at pore volume 13 (fig. 4a) is unknown, but the decrease may have been the result of differences in the kinetics of pyrite oxidation versus calcite dissolution, or a decrease in the availability of calcite particles due to a possible coating of iron oxides and/or hydroxides (products of pyrite oxidation) during the leaching experiments. The pH of all leachates from the mixtures of CSS-1 and FBC residues and mixtures of CSS-1 and sized aglime was approximately 8 (fig. 5). The pH of these leachates was apparently controlled by calcite. In the mixtures, the amount of FBC residues was small, so the conversion of portlandite to calcite was nearly complete, in contrast to the unmixed residues, in which the amount of calcite produced was relatively small compared with the large amounts of portlandite. The data point for the CSS-1 -FBC-4 mixture at 90 days extraction time (fig. 5b) appears to be an outlier. The 180-day extract for this mixture was spilled and was unsalvageable. The pH of leachates from the mixtures of CSS-2 and FBC residues was more variable than the pH of leachates from mixtures of CSS-1 and FBC. Mixtures of CSS-2 and FBC-1, FBC-2, or FBC-5 behaved similarly (fig. 6a). The graph of pH versus time for CSS-2/FBC-4 was similar to those for the other mixtures, but the low-pH portion of the leaching response was delayed, again, possibly because of a coating of iron oxides and/or hydroxides on the calcite particles. The pH of the leachate from the mixture of CSS-2 and FBC-3 at pore volume 3 was acidic, but the pH in the remainder of the leachate samples conformed to the pattern shown by the mixtures of CSS-2 with FBC-1 , FBC-2, and FBC-5. Each leachate from pore volume 15 attained a pH of approximately 8, which suggested a system buffered by calcite. In the batch extraction experiments, the pH of extracts from mixtures of CSS-1 and FBC residues was between 7 and 8 after 30 days of extraction, and the pH remained in that interval for the balance of the extraction periods (fig. 5b). The pH of extracts from mixtures that contained CSS-2 and FBC residue was initially between 12 and 13, but the pH decreased as the extraction period progressed (fig. 6b). The gradual change in pH was indicative of the initial presence of portlandite, followed by its slow conversion to calcite. The slower transition of portlandite to calcite in the batch extraction 15 experiments, compared with the transition of portlandite in the leaching experiments, can be attri- buted to decreased exposure of the portlandite to carbon dioxide. Major Solutes in Leachates and Extracts The predominant solutes in the leachates and extracts from mixtures of CSS and FBC residue or aglime were Ca 2+ , Na + , SCm 2- , and Cl~. Sodium and chloride were quickly flushed during the wet-dry leaching experiments. Their maximum concentrations were generally attained within 3 to 10 days during the batch extractions, owing to their solubility. In comparison with the leachates from the mixtures of CSS and aglime, those from mixtures of CSS and FBC residue had greater Ca 2+ concentrations in the early pore volumes (figs. 7a, 8a). The Ca 2+ concentrations in the last few leachate samples from the mixtures, however, ranged from 400 to 800 mg L" 1 . We attributed the presence of Ca 2+ in the solution to the dissolution of calcite, anhydrite, or gypsum. Sulfate in all the leachates (figs. 9a, 10a) was attributed to the oxidation of pyrite and/or the dissolution of anhydrite. The Ca 2+ and S04 2 " concentrations in the 180-day batch extracts (figs. 7-1 0) were apparently the result of thermodynamic equilibrium with calcite and gypsum (see below). If pyrite oxidation did occur in the batch extractions, the reaction products would have supplied SO4 2 " and Ca 2+ ions to the solution, and the acid produced would have been neutralized by carbonate. The observed changes in the Ca 2+ and SO4 2 " concentrations in the batch extracts were important to understanding the major chemical reactions that occurred in the FBC residue-CSS mixtures. The mean Ca 2+ concentration in all 1 80-day batch extracts was 566 ± 1 8 mg L~ 1 . The Ca 2+ concentrations (fig. 8b) in the extracts from mixtures that contained CSS-2 were initially high because of the solubility of portlandite (1 ,590 mg L -1 at 25°C). With increasing extraction time, the concentration of Ca 2+ in the extracts decreased to a relatively constant value, reflecting the lower solubility of calcite (14 mg L~ 1 at 25°C) when compared with that of portlandite. The possible sources of SO4 2 " in the extracts were the dissolution of anhydrite that was initially present in the solids and the oxidation, if any, of pyrite during the extraction period (figs. 9b, 10b). In general, the concentrations of Ca 2+ and SCm 2- in the leachates and extracts were controlled by the solubility of gypsum, as indicated in the solubility diagrams in figures 11 and 12. In calculating the solubility boundaries, we assumed the gypsum dissolved into a pure aqueous solution. The solutions in this study, however, contained other ions, in particular, sodium and chloride. The presence of other ions increased the ionic strength of the solutions, which increased the solubility of gypsum (and other minerals) relative to its solubility in a solution of low ionic strength. The leachates and extracts were relatively dilute solutions that had ionic strengths of approximately 0.1 (for com- parison, seawater has an ionic strength of approximately 0.7). Ionic interactions, however, would have had some effect on the solubilities of the salts. For example, sulfate ions will interact in solution with cations, such as Ca 2+ , Mg 2+ , and Na + , to form ion pairs, such as CaS04°, MgS04°, and Na2S04°. The concentration of these ion pairs and others would be relatively low when compared with the concentrations of Ca 2+ arid S04 2 ", but they will, nevertheless, increase the solubility of sulfate salts. This ionic strength effect would displace the gypsum solubility boundary upward so that it would lie closer to the points representing the concentration of sulfate on the graphs (figs. 11b, 12b). Con- sideration of the ionic strength effect makes equilibrium with gypsum more reasonable than it appears on the graphs for sulfate ion. As indicated in figures 7 to 10, the concentrations of Ca 2+ and SO4 2 " each converged on or tended to converge on common concentrations for the two ions, another indication of an approach to equilibrium. When the total dissolved concentrations of Ca 2+ , Na + , Cl~, and S04 2- , alkalinity, and pH and Eh values were input to the thermodynamic equilibrium model, MINTEQA2 (U.S. EPA 1990), the calculated results confirmed that after 1 5 pore volumes were leached, the leachates from the wet-dry column experiments would be in equilibrium with both calcite and gypsum. The model was also used to calculate the equilibrated systems for the 180-day extracts. In the case of the batch extracts, however, alkalinity data were lacking. Because the batch solutions were periodically aerated, a constant carbon dioxide partial pressure of 0.0003 atmosphere (the approximate partial pressure of carbon dioxide in air) was used in the model. The modeling results indicated the extracts from the mixtures containing CSS-1 would be in equilibrium with gypsum, and the extracts from the mixtures containing CSS-2 would be in equilibrium with calcite. The latter results are contrary to the results shown by the open triangle symbols for mixtures with CSS-1 (the points between pH 9 and 11) in 16 11 10 X Q. 8 - (a) -•- CSS-1/FBC-1 -•- CSS-1/FBC-2 -A. - CSS-1/FBC-3 -T- CSS-1/FBC-4 ♦ CSS-1/FBC-5 -#•- CSS-1/aglime 1 11 10 I a. 4 6 8 10 12 Pore volumes leached 14 16 9 - 8 - 7 - I I I (b) / / / / / I I I I I -A - CSS-1/FBC-3 -▼- CSS-1/FBC-4 ♦ CSS-1/FBC-5 _ -#•- CSS-1/aglime H I I 30 60 90 120 150 Extraction time (days) 180 Figure 5 pH of leachates and extracts from mixtures of FBC residues and CSS-1 . 14 12 10 1 8 (a) •- CSS-2/FBC-1 •- CSS-2/FBC-2 ■A - CSS-2/FBC-3 ▼- CSS-2/FBC-4 ♦ CSS-2/FBC-5 -• - CSS-2/aglime 14 12 10 i. 8 SSI" \ \ (b) 1 2 4 6 8 10 12 14 16 30 Pore volumes leached Figure 6 pH of leachates and extracts from mixtures of FBC residues and CSS-2. -A - CSS-2/FBC-3 -▼- CSS-2/FBC-4 ♦ CSS-2/FBC-5 -# - CSS-2/aglime 60 90 120 150 Extraction time (days) 180 17 2000 1500 E c o ro 1 000 c CD O c o o CO o 500 (a) CSS-1/FBC-1 CSS-1/FBC-2 CSS-1/FBC-3 CSS-1/FBC-4 CSS-1/FBC-5 - CSS-1/aglime 10 Pore volumes leached 12 14 700 16 c o c CD o c o o CO o 600 - 500 400 30 I I h r- hi -»♦ 1 (b) 1 1 1 1 1 I -A- CSS-1/FBC-3 -T- CSS-1/FBC-4 •♦• CSS-1/FBC-5 -#■- CSS-1/aglime I I 60 90 120 Extraction time (days) 150 180 Figure 7 Calcium concentration in leachates and extracts from mixtures of FBC residues and CSS-1 . o ro c CD O c o o CO o 2000 _ 1500 - D) E 1000 - 500 I I I I I I CSS-2/FBC-1 -«- CSS-2/FBC-2 -A- CSS-2/FBC-3 -▼- CSS-2/FBC-4 _ j\ \ ' \ CSS-2/FBC-5 CSS-2/aglime — — *v *^ ^\ — ^Zi^^A — ^^3^ X-#-"" (a) l l l l I I 2000 - _ 1 500 E V c o CO h_ •4-* c CD o c o o CO o 1000 500 -A - CSS-2/FBC-3 -T- CSS-2/FBC-4 ♦ CSS-2/FBC-5 -#•- CSS-2/aglime V \ \ \ \ --•^^^ (b) 1 2 4 6 8 10 12 14 16 30 60 90 120 Pore volumes leached Extraction time (days) Figure 8 Calcium concentration in leachates and extracts from mixtures of FBC residues and CSS-2. 150 180 18 E c o O CO HUUU ~T 1 1 i i i i -#- CSS-1/FBC-1 -fh CSS-1/FBC-2 3500 1 -A - CSS-1/FBC-3 - -T- CSS-1/FBC-4 \ \ ♦ CSS-1/FBC-5 -#■- CSS-1/aglime 3000 — — \ 2500 r ^ 2000 ^ ^r^~ tc.r\r\ (a) 1 i i I I I I o CO 2800 2400 - 2000 O) E c o iz 1600 c 0) o c o o 1200 800 400 I T fits I I I -^1 I / -I / - I/ ^. CSS-1/FBC-3 -T- CSS-1/FBC-4 ♦ CSS-1/FBC-5 (b) I I I -#■- CSS-1/aglime I 4 6 8 10 12 Pore volumes leached 14 16 30 60 90 120 Extraction time (days) Figure 9 Sulfate concentration in leachates and extracts from mixtures of FBC residues and CSS-1 . 150 180 E c o O CO 5000 4500 4000 3500 - 3000 2500 2000 1500 T T T T (a) CSS-27FBC-1 CSS-2/FBC-2 CSS-2/FBC-3 CSS-2/FBC-4 CSS-27FBC-5 CSS-2/aglime 1 ± 4 6 8 10 12 Pore volumes leached 14 16 E c o O CO 3000 2500 - 1500 1000 500 I . : 7 I I / I I zzzz^ ♦/ ^4 -A - CSS-2/FBC-3 -"T- CSS-27FBC-4 ♦ CSS-2/FBC-5 (b) I I I -#•- CSS-2/aglime I I 30 60 90 120 Extraction time (days) 150 Figure 10 Sulfate concentration in leachates and extracts from mixtures of FBC residues and CSS-2. MAY 1 K 1007 180 19 + CM (0 O D) O -2 - -3 1 1 \ \ o i i \ 2- Supersaturated \% — \ w l 2- \ c \ -• \8 — •• • r**rf — £ _ _u- ( Gypsum solubility— Undersatu rated (a) l 1 1 1 1 I CM o CO o -1 - -2 - -3 I I I I Supersaturated I J* W ^W * • • • •• — Gypsum solubility — Undersaturated (b) I I I I I 10 12 10 12 PH PH Figure 11 Molar concentrations of calcium and sulfate versus pH in leachates from mixtures of FBC residues and CSS. The filled circles represent leachates from pore volumes 3 through 13. The open triangles represent leachates from pore volume 15. The lines representing the solubilities of calcite and gypsum were calculated from data by Dean (1985). CO O D) O -2 - -3 I \ I \ ° 1 1 \ 2» \ 2 Supersaturated — \ o \ c \ rs. — • .# • — •■Sf" 1 r% — za — -*-&m — — — Gypsum solubility Undersaturated! (a) I 1 1 I CM o CO D) O -1 -2 - I I I Supersaturated I _»^P^ m _ ^ Gypsum solubility Undersaturated (b) I I I I 10 12 10 12 PH PH Figure 12 Molar concentrations of calcium and sulfate versus pH in extracts from mixtures of FBC residues and CSS. The filled circles represent extracts from 3 through 90 days extraction time. The open triangles represent extracts from 180 days. The lines representing the solubilities of calcite and gypsum were calculated from data by Dean (1985). 20 Table 10 State of Illinois General Use Water Quality Standards and ranges of concentrations of various inorganic chemicals in leachates and extracts from mixtures of CSS and FBC residues. Range of concentrations in leachates and extracts Illinois Standard* (mg L" 1 ) from mixtures (mg L" 1 ) Constituent CSS-1 CSS-2 PH 6.5-9.0 6.77-8.79 2.86-12.6 Ag 0.005 t As 0.36 <0.001-0.04 <0.001-0.08 B 5.0 1.7-12 0.03-20 Ba 5.0 0.01-0.07 0.01-0.20 Cd 0.05 <0.01 <0.01 CI 500 4.21-5243 12.0-2598 Cr* 31 <0.01-0.06 <0.01-0.05 Cu 0.5 <0.01-0.01 <0.01-0.25 F 1.4 1.00-11.7 0.99-177 Fe 1.0 <0.01-0.03 <0.01-2.08 Hg 0.0005 <0.02-0.03 <0.02-0.03 Mn 1.0 <0.004-0.83 <0.004-13.5 Ni 1.0 <0.03-0.07 <0.03-2.90 Pb 0.1 <0.08 <0.08 Se 1.0 <0.01-1.4 <0.01-3.3 S0 4 2 " 500 840-4211 1030-4746 Zn 1.0 <0.004-0.057 <0.004-0.12 'From State of Illinois Rules and Regulations (1995) f — = not determined figure 12a. The contradiction between the modeling results and the graphical results might have been caused by the use of carbon dioxide partial pressure in modeling rather than alkalinity values. Table 10 compares the concentrations of various constituents in the leachates and extracts from mixtures of FBC residues and CSS with the Illinois General Use Water Quality Standards (State of Illinois 1995). Several constituents were present at concentrations below the respective standard in all leachates and extracts, but most were present at concentrations above the respective standard in some of the leachates and/or extracts. Only SO4 2 " was present at concentrations above its standard in all leachates and extracts. The leachates from mixtures of CSS and FBC residue or aglime had higher concentrations of both Na + and CI" than those from unmixed samples. At pore volumes 9 through 15, Na + concentrations were 700 mg L~ 1 or less and CI" concentrations were less than 400 mg L~ 1 in leachates from mixtures of CSS-2 and FBC residues. In all other leachates, the concentrations of Na + were 200 mg L" 1 or less, and CI" concentrations were 65 mg L _1 or less. At low concentrations, chloride is an essential nutrient for plants, but at 500 mg L~ 1 CI" is toxic to most sensitive plants (Severson and Shacklette 1988), and it would probably be toxic to plants used in revegetation (Gough et al. 1979). Excess CI" will cause yellowing and burning in leaves. Sodium interferes with a plant's ability to absorb water through the roots. In addition, sodium required five to seven pore volumes of leaching to decrease its concentration to less than 500 mg L~ 1 . The presence of these ions in the pore water of a codisposed mixture of FBC residue and CSS is a cause for concern and will need to be monitored during the early phases of revegetation. The concentration of K + was high (up to 2700 mg L _1 ) in leachates from the unmixed samples of FBC-1, FBC-2, and FBC-5, whereas in the extracts, the highest concentration was 220 mg L~ 1 . The concentration of Mg 2+ was generally higher in leachates from mixtures containing CSS-2 (<230 mg L~ 1 ) than in those containing CSS-1 (<70 mg L~ 1 ), and leachates from mixtures that contained aglime had higher concentrations of Mg 2+ than those that contained FBC residues. In the batch extracts, the Mg 2+ concentrations from mixtures that contained CSS-1 were higher than in those from mixtures that had CSS-2. The unmixed CSS-2 contained twice the concentration of Mg (0.8% as MgO) as the unmixed CSS-1 (0.4% as MgO). Unmixed FBC residues contained concen- 21 trations of Mg (as MgO) between 0.62% and 1 .05%. Neither K + nor Mg 2+ appeared to be present in the leachates or extracts of the mixtures at concentrations that would be toxic to plants. The trace elements of note in the leachates from mixtures with either CSS-1 or CSS-2 were B, Se, Ni, and Mo (table 10). Boron concentrations were independent of pH and were higher in leachates from the unmixed aglime sample (0.20-0.36 mg L~ 1 ) than in those from the unmixed FBC residues (0.03-0.18 mg L~ 1 ), but they were highest in the leachates from the unmixed CSS materials (0.66-3.9 mg L _1 ). The concentration of B in the extracts ranged from 0.03 to 3.2 mg L -1 . Severson and Shacklette (1988) indicated that sensitive plants may show symptoms of toxicity to B when the concentration in water used for irrigation is greater than 0.75 mg L~ 1 . Boron may be removed by adsorption on AKOH3), and the liming of acidic soils increases the adsorption. Boron in the codisposal situation could quite possibly occur at subtoxic concentrations. Selenium and nickel occurred at elevated concentrations in leachates from mixtures with CSS-2. Selenium, which occurs as an anion in aqueous solution, was mildly correlated with pH (r 2 = 0.68) in leachates from CSS-1 mixtures. Nickel was correlated with other chalcophile elements, such as selenium and copper, but not with SO4 2 ". These findings for nickel may indicate that pyrite oxidation was not the sole source of sulfate in the leachates, but that anhydrite was also a contributor. Molybdenum, which also occurs as an anionic species in aqueous solution, was moderately correlated with pH (r 2 = 0.77). For those solutes that displayed a correlation with pH, one can expect that the concentration of such a solute will change if the pH of the leachate changes in a codisposal situation. The concentration of F~ in the leachates after about the seventh pore volume generally ranged from 2 to 4 mg L~ 1 . The concentration of F - in the extracts ranged from 1 to 5 mg L _1 . If the codisposed mixture of CSS and FBC residue is not allowed to leach prior to sowing seeds, toxic concentrations of F _ could be present. Manganese is not expected to be toxic to plants used in revegetation, especially if the growth medium is maintained in a slightly alkaline condition. Of the various aqueous species of Mn, only Mn 2+ is absorbed by plants (Severson and Shacklette 1988). At alkaline pH, Mn is effectively removed from the soil solution by precipitation as oxyhydroxides. Selenium was found to substitute for S in pyrite (Finkelman 1981). If Se was associated with the pyrite in the CSS materials, then as pyrite oxidation proceeded, Se would have been released to the leachate solution. Selenium in the FBC residues was probably associated with the coal ash portion of the residues. The greatest concentrations of Se (fig. 1 3) were detected in leachates from mixtures with CSS-2, which had the greater pyrite content of the two coal slurry samples. Selenium attained steady-state concentrations of <1 mg L~ 1 in mixtures containing CSS-2 and <0.05 mg L~ 1 in the later pore volumes from mixtures containing CSS-1 . Sulfate is probably the only form of sulfur absorbed by plants (Severson and Shacklette 1988), and it is used in metabolic functions, such as in the production of chlorophyll and the synthesis of amino acids. The presence of SO4 2 " at the observed concentrations in the pore water of a codisposed mixture of FBC residue and CSS is not expected to be deleterious to the plants used in reclamation of the site (Gough et al. 1979). The concentrations of Al, As, Be, Cd, Cr, Cu, Hg, Mo, Se, and Zn in the leachates were less than the levels indicated by Gough et al. (1979) as being toxic to plants. B, Cl~ Co, F~ Mn, and Ni at their highest observed concentrations could be toxic to some plants, depending on the sensitivity of the plant to the element and the ability of the plant to absorb that element. The latter constituents were present at potentially toxic concentrations in some leachates from some solid samples, and in these cases, the mixture usually contained CSS-2. Sodium, another element potentially toxic to plants, was present in the initial leachates at concentrations that might inhibit the ability of plants to absorb water through the roots. SUMMARY The particle-size distribution of FBC residues varies between residues. Those residues with a predominance of smaller particles (FBC-1, FBC-2, and FBC-5) are expected to dissolve more quickly than the others upon weathering. Therefore, application of an FBC residue having small particles to CSS may be required more frequently than those with larger particles. The use of only bed ash rather than a mixture of fly ash and bed ash may be advisable in codisposal because bed ash particles are typically about 1 to 2 mm or larger in diameter. 22 0.7 0.6 0.5 !_, 0.4 O) E £ 0.3 'c Q) «> 0.2 0.1 0.0 i r -#- CSS-1/FBC-1 m- CSS-1/FBC-2 -A- CSS-1/FBC-3 ■T- CSS-1/FBC-4 ♦ CSS-1/FBC-5 ■#•- CSS-1/aglime (a) 3.5 3.0 2.5 D) 70 E *■— <* E ^ c 1.5