OFR-61-92 An Assessment Of Factors Affecting Pyrite Reactivity

Variations in pyrite reactivity are known to exist within samples collected from different areas. Using samples collected from acid and non-acid sites, the relationship between the factors affecting pyrite oxidation and observed variations in pyrite reactivity was studied. Heavy liquid separations were used to extract the pyrite from nine pulverized rock, coal, and coal refuse samples. Simulated weathering tests were used to determine the degree of reactivity of the extracted pyritic material. Representative portions of each pyritic sample were analyzed for whole rock composition (including pyritic sulfur and neutralization potential), for elemental composition, and by x-ray diffractometry. The samples were also examined with reflected light microscopy to ascertain pyrite morphology distributions. Tests for the presence of iron-oxidizing bacteria were all negative indicating the leaching tests were conducted under abiotic conditions. The results of the simulated weathering tests permitted the partitioning of the samples into two groups based on positive or negative neutralization potential (NP) values. Determinations were then made to evaluate the degree to which a variety of factors, including temperature, partial pressure of oxygen and water, and crystallography affected pyrite reactivity. Multiple regression analyses showed no statistically significant relationship between degree of reactivity and elemental composition. However, for samples with negative NPs, a simple regression analysis showed a strong linear relationship to exist between the reactivity of the pyrite and the proportion of "unstable" pyrite (framboidal + mossy pitted) found in each extracted pyrite sample. Samples of pyrite with positive NPs, although producing acidity, did so in at least one order of magnitude less than the NP deficient samples, indicating the important role that NP plays in controlling pyrite reactivity. A simple linear regression between cumulative sulfate production and NP for the three samples with positive NPs had a correlation coefficient of 0.99. For the nine samples grouped together, a linear regression model between cumulative sulfate production and NP (both negative and positive) was established between log of sulfate and NP.