EFFECTS OF FLY ASH ON MERCURY OXIDATION DURING POST COMBUSTION CONDITIONS

Tests were performed in simulated flue gas streams using fly ash from the electrostatic precipitators of two full-scale utility boilers. One fly ash was from a Powder River Basin (PRB) coal, while the other was from Blacksville coal. Elemental Hg was injected upstream from samples of fly ash loaded onto filters housed in an oven at 120 or 180 C. Concentrations of oxidized and elemental Hg downstream from the filters were determined using the Ontario Hydro method. The gas stream composition and whether or not ash was present in the gas stream were the two most important variables affecting Hg oxidation. The presence of HCl, NO, NO{sub 2}, and SO{sub 2} were all important with respect to Hg oxidation, with NO{sub 2} and HCl being the most important. The presence of NO suppressed Hg oxidation in these tests. Although the two fly ashes were chemically and mineralogically diverse, there were generally no large differences in catalytic potential (for oxidizing Hg) between them. Similarly, no ash fraction appeared to be highly catalytic relative to other ash fractions. This includes fractions enriched in unburned carbon and fractions enriched in iron oxides. Although some differences of lesser magnitude were observed in the amount of oxidized Hg formed, levels of oxidized Hg generally tracked well with the surface areas of the different ashes and ash fractions. Therefore, although the Blacksville fly ash tended to show slightly more catalytic activity than the PRB fly ash, this could be due to the relatively high surface area of that ash. Similarly, for Blacksville fly ash, using nonmagnetic ash resulted in more Hg oxidation than using magnetic ash, but this again tracked well with the relative surface areas of the two ash fractions. Test results suggest that the gas matrix may be more important in Hg oxidation chemistry than the fly ash composition. Combustion tests were performed in which Blacksville and PRB fly ashes were injected into filtered (via a baghouse with Teflon bags) flue gas obtained while firing PRB coal in a 35 kW combustor. The Ontario Hydro method was used to determine the Hg speciation after fly ash injection. Wall effects in the combustor complicated interpretation of testing data, although a number of observations could still be made. The amount of Hg collected in the Ontario Hydro impingers was lower than anticipated, and is probably due to sorption of Hg by the fly ash. While firing PRB coal without any ash injection, the percent oxidized Hg in the gas stream was fairly high (average of 63%). The high levels of vapor phase oxidized Hg in these base line tests may be due to catalytic effects from the refractory materials in the combustor. When PRB fly ash was injected into a filtered PRB flue gas stream, the percentage of oxidized Hg in the gas stream decreased dramatically. Decreases in the percentage of oxidized Hg were also observed while injecting Blacksville fly ash, but to a lesser extent. Injecting whole Blacksville fly ash into the filtered PRB flue gas appeared to result in greater concentrations of oxidized Hg relative to the tests where whole PRB fly ash was injected. However, because the Blacksville fly ash has a relatively high surface area, this may be only a surface area effect