Mode of occurrence analyses indicate that trace elements (Sb, As, Cd, Cr, Co, Cu, Pb, Hg, Ni, Se, V, and Zn) in the Illinois No. 6 coal are generally associated with relatively large discrete mineral grains, especially pyrite, whereas trace elements in Absaloka coal are much more strongly associated with macerals and fine-grained minerals. These coals were burned using conventional and low-NO x conditions in an ∼7-kW combustion system to evaluate the importance of elemental modes of occurrence, coal properties, and combustion conditions on trace element volatility and As, Cr, and Ni speciation. Chemical analyses of size-classified (∼0.4-7.7 µm) fly ash and flue gas samples indicated that Hg and Se were the most volatile elements in both coals. Occurrences of Cr, Co, and Cu in fly ash were characterized by relatively uniform particle-size distributions and relative enrichment/depletion (RED) factors for all four fly ashes, which is indicative of nonvolatility. As and Sb in Absaloka fly ashes exhibited similar nonvolatile partitioning characteristics. Consistent with an elemental vaporization-particle surface deposition process, Sb, As, Cd, Pb, Ni, V, and Zn concentrations and RED factors for the Illinois No. 6 fly ashes generally increased with decreasing particle size. Similar semivolatile partitioning systematics were noted for Cd, Pb, and V in Absaloka fly ashes. Conventional and low-NO x combustion of Illinois No. 6 coal did not significantly affect trace element volatility. However, low-NO x Absaloka combustion promoted Ni, Zn, and Se volatilization. The inorganic phase composition and As, Cr, and Ni speciation of fly ash particles ∼2.5 µm in aerodynamic diameter (FA 2.5 ) were determined using X-ray diffraction and absorption methods. Illinois No. 6 and Absaloka FA 2.5 contain aluminosilicate glass, quartz (SiO 2 ), ferrite spinel (AB 2 O 4 ; e.g., where A 2+ ) Fe, Mg, Ni, Co, Cu and B 3+ ) Al, Fe, Cr), and mullite (Al 6 Si 2 O 13 ). Absaloka FA 2.5 is distinguished from Illinois No. 6 FA 2.5 by the presence of lime (CaO) and periclase (MgO) and lack of anhydrite (CaSO 4 ). Differences in Illinois No. 6 and Absaloka coal combustion conditions did not significantly affect As, Cr, or Ni speciation. As 5+ O 4 -containing phases occur in Illinois No. 6 and Absaloka FA 2.5 . Presumably, carboxyl-bound As 3+ and Ca in Absaloka coal promoted the formation of Ca 3 (AsO 4 ) 2 . Cr 3+ /Cr 6+ is much greater in Illinois No. 6 FA 2.5 , relative to Absaloka FA 2.5 . The predominance of maceral-bound Cr 3+ and oxygen functional groups in Absaloka coal may have promoted Cr 6+ formation. Illinois No. 6 and Absaloka FA 2.5 contain similar NiO-bearing phases, possibly ferrite spinel.

The chemical speciation of Ni in fly ash produced from approximately 0.85 wt % S residual (no. 6 fuel) oils in laboratory (7 kW)- and utility (400 MW)-scale combustion systems was investigated using X-ray absorption fine structure (XAFS) spectroscopy, X-ray diffraction (XRD), and acetate extraction [1 M NaOAc-0.5 M HOAc (pH 5) at 25 degrees C]-anodic stripping voltammetry (ASV). XAFS was also used to determine the Ni speciation of ambient particulate matter (PM) sampled near the 400-MW system. Based on XAFS analyses of bulk fly ash and their corresponding acetate extraction residue, it is estimated that > 99% of the total Ni (0.38 wt %) in the experimentally produced fly ash occurs as NiSO4.xH2O, whereas > 95% of the total Ni (1.70 and 2.25 wt %) in two fly ash samples from the 400-MW system occurs as NiSO4.xH2O and Ni-bearing spinel, possibly NiFe2O4. Spinel was also detected using XRD. Acetate extracts most of the NiSO4.xH2O and concentrates insoluble NiFe2O4 in extraction residue. Similar to fly ash, ambient PM contains NiSO4.xH2O and NiFe2O4; however, the proportion of NiSO4.xH2O relative to NiFe2O4 is much greater in the PM. Results from this and previous investigations indicate that residual oil ash produced in the 7-kW combustion system lack insoluble Ni (e.g., NiFe2O4) but are enriched in soluble NiSO4.xH2O relative to fly ash from utility-scale systems. This difference in Ni speciation is most likely related to the lack of additive [e.g., Mg(OH)2] injection and residence time in the 7-kW combustion system.

Ash from a low-and high-S (0.33 wt% and 1.80wt% S, respectively) residual oil was produced using a laboratory-scale combustion system at excess O 2 concentrations of $ I and 2 or 3 mol%. High-S ashes are distinguished from low-S ashes by an abundance of (Na, K)x v~+ vtx 0" (0.90> x > 0.54) and lack of (Na, Kj,SO. crystals. Discrete phases of Ni or Cr were not detected using SEM and XRD, even though these metals are relatively abundant -\.5 to 5.5 wt% and 0.08 to 0.1 wt%, respectively. Ni and Cr K-edge XAFS s~ectroscopy analyses indicate that NiSO. and Cr2(SO.h and not the more toxic Ni,S2 and Cr· forms predominate in the ashes. Thermodynamic modeling results support the empirical results in that NiSO•. XH 2 0 and Cr2(SO.h are predicted to be stable low-temperature species in both lowand high-S residual oil ashes produced at S 3 mol% excess O 2,

No abstract

Representative duplicate fly ash samples were obtained from the stacks of 400- and 385-MW utility boilers (Unit A and Unit B, respectively) using a modified U.S. Environmental Protection Agency (EPA) Method 17 sampling train assembly as they burned 0.9 and 0.3 wt % S residual (No. 6 fuel) oils, respectively, during routine power plant operations. Residual oil fly ash (ROFA) samples were analyzed for Ni concentrations and speciation using inductively coupled plasma-atomic emission spectroscopy, X-ray absorption fine structure (XAFS) spectroscopy, and X-ray diffraction (XRD). ROFA deionized H2O extraction residues were also analyzed for Ni speciation using XAFS and XRD. Total Ni concentrations in the ROFAs were similar, ranging from 1.3-1.5 wt%; however, stack gas Ni concentrations in the Unit A were 0.990 microg/Nm3 compared with 0.620 microg/Nm3 for Unit B because of the greater residual oil feed rates employed at Unit A to attain higher 400-MW load conditions with a lower heating value oil. Ni speciation analysis results indicated that ROFAs from Unit A contain approximately 3 wt % NiSO4 x xH2O (where x is assumed to be 6 for calculation purposes) and appoximately 4.5 wt% of a Ni-containing spinel compound, similar in composition to (Mg,Ni)(Al,Fe)2O4. ROFAs from Unit B contain on average 2 wt% NiSO4 x 6 H20 and 1.1 wt% NiO. XAFS and XRD analyses did not detect any nickel sulfide compounds, including carcinogenic nickel subsulfide (Ni3S2) (XAFS detection limit is 5% of the total Ni concentration). In addition, XAFS measurements indicated that inorganic sulfate and organic thiophene species accounted for > 97% of the total S in the ROFAs. Unit A ROFAs contained much lower thiophene proportions because cyclone-separated ROFA reinjection is employed on this unit to collect and reburn the larger carbonaceous particles.