Metal Release and Speciation Changes during Wet Aging of Coal

Nelson Rivera , Dean Hesterberg , Navdeep Kaur , and Owen W. Duckworth ... Grace E. Schwartz , James C. Hower , Allison L. Phillips , Nelson Rivera , ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/est

Metal Release and Speciation Changes during Wet Aging of Coal Fly Ashes Jeffrey G. Catalano,†,* Brittany L. Huhmann,†,§ Yun Luo,†,‡,∥ Elizabeth H. Mitnick,† Adam Slavney,† and Daniel E. Giammar‡ †

Department of Earth and Planetary Sciences, Washington University, Saint Louis, Missouri 63130, United States Department of Energy, Environmental, and Chemical Engineering, Washington University, Saint Louis, Missouri 63130, United States



S Supporting Information *

ABSTRACT: Introduction of coal fly ash into aquatic systems poses a potential environmental hazard because of its heavy metal content. Here we investigate the relationship between solid phase transformations, fluid composition, and metal release and speciation during prolonged wet aging of a class C and class F coal fly ash. The class C ash causes rapid alkalinization of water that is neutralized over time by CO2 uptake from air and calcite precipitation. The resulting aqueous metal concentrations are below regulatory limits with the exception of Cr; solubility constraints suggest this is released as chromate. Limited As release is accompanied by no change in solid-phase speciation, but up to 35% of the Zn in the ash dissolves and reprecipitates in secondary phases. Similar processes inhibit Ba and Cu release. In contrast, the class F ash causes rapid acidification of water and initially releases substantial quantities of As, Se, Cr, Cu, Zn, and Ba. Arsenic concentrations decline during aging because of adsorption to the iron oxide-rich ash; this is aided by As(III) oxidation. Precipitation processes lower Ba and Cr concentrations during aging. Se, Cu, and Zn concentrations remain elevated during wet aging and solid-phase Zn speciation is not affected by ash-water reactions. Total metal contents were poor predictors of metal release, which is predominantly controlled by metal speciation and the effects of ash-water reactions on fluid pH. While contact with atmospheric gases has little effect on class F ash, carbonation of class C ash inhibits metal release and neutralizes the alkalinity produced by the ash.



INTRODUCTION Fly ash is a major byproduct of coal combustion for energy generation, with 61.4 million metric tons produced in 2010 in the United States.1 While fly ash is reused for various purposes, including cement production, approximately 60% is disposed of in ash ponds and landfills.1 Although currently exempt from regulation as hazardous waste in the U.S., fly ash contains elevated concentrations of heavy metals2 and potentially poses an environmental hazard if introduced into aquatic systems. Following the recent large release of coal fly ash into the Emory River as a result of the structural failure of an ash storage pond in Kingston, TN,3 the U.S. Environmental Protection Agency has proposed new regulations to manage coal ash disposal.4 Fly ash may undergo extensive alteration upon exposure to water, including dissolution and reprecipitation and release of heavy metals.2,5,6 Solid dissolution processes upon exposure to water vary with the type of fly ash, with CaO-rich class C ash7 generally producing alkaline fluids5 and CaO-poor class F ash7 producing acidic fluids.5 In addition, these ash types differ in the relative amounts of major elements available to generate secondary solid phases.6 Leaching of heavy metals from fly ashes is strongly affected by the speciation of such elements in the ash8−18 and the nature of ash-water reactions.2,19 However, © 2012 American Chemical Society

it is unclear how metal release from and metal speciation in coal fly ash are affected by solid phase transformations associated with ash-water reactions during prolonged wet aging. In this study we investigated the relationship between solid phase transformations, fluid composition, and metal release and speciation during wet aging of coal fly ashes. The behaviors of class C and class F ash aged in simulated rainwater with periodic controlled aeration were compared. The experimental conditions were designed to approximate conditions experienced by ashes exposed to meteoric waters. The evolution of the fluid and solid phases were monitored during the aging process. The speciation of arsenic and zinc during aging were characterized using X-ray absorption fine structure (XAFS) spectroscopy. These elements were selected for detailed study as examples of anionic and cationic contaminants. In addition, these elements occur in the ashes in solid phase concentrations amenable to study by XAFS spectroscopy and their initial speciation in the class C ash was previously reported.16 Received: Revised: Accepted: Published: 11804

July 11, 2012 October 1, 2012 October 4, 2012 October 4, 2012 dx.doi.org/10.1021/es302807b | Environ. Sci. Technol. 2012, 46, 11804−11812

Environmental Science & Technology

Article

Figure 1. Powder XRD patterns of the (A) class C and (B) class F fly ash solids as a function of aging time. Select major diffraction lines for the predominant crystalline components are indicated: A = anhydrite [CaSO4], C = calcite [CaCO3], G = gehlenite [CaAl(AlSiO7)], H = hematite [αFe2O3], L = lime [CaO], Me = merwinite [Ca3Mg(SiO4)2], Mh = maghemite [γ-Fe2O3], Mu = mullite [nominally Al6Si2O13], P = periclase [MgO], Q = quartz [SiO2]. A full tabulation of strong diffraction lines for these phases is provided in SI Table S3.



MATERIALS AND METHODS Ash Samples. Class C and F fly ash samples were obtained from electric power utility companies in eastern Missouri and southwestern Indiana, respectively. The class C ash was produced at a plant burning subbituminous coal from the Powder River Basin and has been previously described16 (ash sample B). The class F ash was produced at a plant burning bituminous coal obtained from the region surrounding the plant. Aging Procedure. To simulate fly ash aging in the environment, ash samples were reacted in batch reactors with synthetic rainwater having the composition of rain in eastern Missouri.20 The synthetic rainwater was prepared from reagentgrade (NH4)2SO4, Ca(NO3)2, Na2SO4, MgCl2, and KCl, and by dilution of 1.0 M standards of HCl, HNO3, and H2SO4. The final solution contained 0.078 mg L−1 Na+, 0.027 mg L−1 Mg2+, 0.030 mg L−1 K+, 0.27 mg L−1 Ca2+, 0.32 mg L−1 NH4+, 1.31 mg L−1 SO42‑, 0.86 mg L−1 NO3−, and 0.094 mg L−1 Cl− and had a measured pH of 5.04 after equilibration with atmospheric CO2. For each reactor 0.1 g of ash was added to 10 mL of synthetic rainwater; 60 replicates were prepared for each ash type. Samples were continuously mixed on end-over-end rotators and were removed once per week for pH measurements and sparging with humidified air for 15 min in order to provide controlled introduction of atmospheric CO2 and O2. pH was measured with a VWR Ag/AgCl electrode and a Thermo Scientific Orion meter. The electrode was calibrated weekly at pH 4.01, 7.00, and 10.01 using fresh buffers. pH

measurements for 1 week and longer represent the average of the set of samples existing at the time of measurement (e.g., 60 samples at week 1, 30 samples at week 31); reported error bars are 1 standard deviation of the set of pH measurements for the sampling period. Once per week a single sample was collected for aqueous and solid phase analysis prior to air sparging. The sample was centrifuged at 4000g for 10 min and the liquid phase was decanted, filtered through a 0.22 μm syringe filter (MCE), and acidified with HNO3 for preservation; final pH was measured before centrifugation. The solid was then washed three times by resuspending in deionized water (>18.2 MΩ·cm), centrifuging, and decanting the supernatant to remove residual dissolved salt. The washed ash was then dried at 70 °C overnight prior to solid-phase analysis. Additional samples were prepared with aging times ranging between 15 min and 1 week following the same procedures; only the supernatant was analyzed. Fluid Characterization. Fluid compositions were determined using an Agilent 7500ce inductively coupled plasma mass spectrometer (ICP-MS). All samples were diluted 1:5 and 1:1000 in 2% HNO3 to bring the elements of interest into the working range of the instrument and to reduce the total dissolved solids. The minimum detectable concentrations (MDCs), which are the method detection limits scaled by the dilution factor required for each element of interest, were specific to each ash experiment (Supporting Information (SI), Table S1) because of different matrix effects caused by ash dissolution and variations in instrument performance over time. 11805

dx.doi.org/10.1021/es302807b | Environ. Sci. Technol. 2012, 46, 11804−11812

Environmental Science & Technology

Article

Figure 2. Evolution of fluid pH and composition during wet aging of the class C (left) and class F (right) coal fly ashes. The open symbol is the pH of the initial synthetic rainwater. Values in parentheses following element labels indicate that the concentration of that element has been rescaled by the value listed to facilitate plotting on a single axis.

positions and intensities to those calculated from the crystal structures of relevant phases (see SI section S1). X-ray absorption fine structure (XAFS) spectra of unreacted and wet aged ashes were measured at the Advanced Photon Source (APS) beamlines 20-BM-B (PNC/XSD) and 5-BM-D (DND-CAT). Both beamlines employed Si (111) doublecrystal monochromators to select the incident beam energy; the second crystal was detuned 10 to 30% (depending on the absorption edge and beamline) to reduce the harmonic content of the beam. Both beamlines also employed mirror systems for harmonic rejection and beam optimization. The configuration at 20-BM-B was previously described.16 At 5-BM-D, a pair of Rh-coated Si mirrors was used for collimation and to reduce the harmonic content of the beam. Dried samples were loaded into polycarbonate sample holders and sealed with Kapton tape for XAFS analysis. XAFS spectra, including both the X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) regions, were measured at the Zn and As K-edges in fluorescence yield using either a 13-element energydispersive Ge solid state detector (20-BM-B and some 5-BM-D measurements) or a pair of four-element energy-dispersive Si

The thermodynamic favorability of select precipitation and redox reactions was examined in The Geochemist’s Workbench21 using previously described procedures and database contents.22 Revised thermodynamic data for Cr(III) hydrolysis and precipitation were added to the database.23 Solid Characterization. The initial ash samples were characterized for major and trace element composition by X-ray fluorescence (XRF) using previously described procedures.24,25 Total C and S contents were determined using a Costech ECS 4010 Elemental Analyzer. Powder X-ray diffraction (XRD) patterns were collected on the initial ashes and a subset of the reacted solids using a Rigaku DMAX/A diffractometer using Cu Kα radiation equipped with a graphite crystal monochromator before the detector to suppress sample fluorescence. Patterns were measured from 15 to 45° 2θ with a step size of 0.04° and a counting time of 4 s per step. The only data processing was the subtraction of the low-angle background, which was modeled as the tail of a Gaussian distribution centered at 2θ = 0°. This was done to preserve the broad feature in the middle of each pattern caused by scattering from the glass component of each sample. Phase identification was performed by comparing peak 11806

dx.doi.org/10.1021/es302807b | Environ. Sci. Technol. 2012, 46, 11804−11812

Environmental Science & Technology

Article

Figure 3. As (A−C) and Zn (D−F) K-edge XANES spectra (A,D), EXAFS spectra (B,E), and magnitudes of the Fourier transforms of the EXAFS spectra (C,F) of the class C fly ash as a function of aging time. For the EXAFS spectra and corresponding Fourier transforms, both data (dotted) and fits (solid) are shown.

2). The class C ash raises the pH to 11 within 15 min; this plateaus at pH 11.4 after 1 week. After 4 weeks of aging time base generation by the ash slows and CO2 uptake begins to lower the pH. After 25 weeks the pH levels off at 9.5; the two low pH measurements at weeks 25 and 26 are likely an electrode error. In the first week Ca shows rapid release to 230 mg L−1 and then drops to 120 mg L−1; it then varies between 75 and 120 mg L−1 through the end of the study. Al concentrations quickly reach 40−50 mg L−1 and stay constant for 20 weeks before declining to