Data from analysis of mineralogy, water extracts, and groundwater demonstrate the significance of the neutralization of the acid from the tailings by the carbonate in the soil on the processes of retardation of contaminants. The observed data are consistent with the results of the thermodynamic calculations done with PHREEQE for the concentrations of components exiting the reaction zone.
X-ray diffraction data from the Salt Lake City tailings illustrate the effectiveness of calcite in the soil in buffering the pH of the groundwater (Figure 2). Gypsum is the dominant calcium-bearing mineral within the tailings; whereas, calcite is dominant in the soils. Below the base of the tailings in an interface zone of less than 10 cm in depth, calcite coexists with gypsum. This suggests that some sulfuric acid from the tailings has penetrated the soil and caused the replacement of some calcite by gypsum. The pH increased from 4.9 to 7.4 across this interface zone. Mineralogic analysis of the soil below the interface zone shows unaltered calcite and the absence of gypsum which suggests the lack of seepage from the tailings.
The observed mineralogical sequence is consistent with the chemistry of water and acid extracts of samples of cores through the tailings and into the subjacent soil. Figure 3 shows the vertical trends of water- and acid-soluble material in a core through the Riverton tailings. The outside profile is the sum of acid- and water-soluble material, and the inside line is the water-soluble portion. At the interface between tailings and soil, the pH increases and Eh decreases. Sulfate is mostly water soluble in both the tailings and soils. Aluminum, iron, manganese, calcium, magnesium, and the trace elements are primarily water soluble in the tailings and in the soils as acid-soluble phases, probably hydroxides and carbonates. Of the trace metals in the soil extracts, arsenic, cadmium, chromium, and vanadium are completely in the acid-soluble phases; molybdenum is predominantly in the water-soluble phases; and nickel and uranium are partitioned between the acid- and water-soluble phases.
The chemical reactions at the interface and the accumulation of elements appear to have developed primarily during the time when the tailings were being used. During this time, tailings liquor was ponded on the tailings, creating a hydraulic head. Drainage occurred throughout the active life of the tailings, and after the tailings became inactive until hvdrologic equilibrium was established. Continued seepage due to infiltration from ruin and show appears to have had minimal effects on the leaking of most contaminants from the tailings.
The trends of decreasing concentrations in Figure 3 define a disequilibrium zone of about one meter. This suggests that all of the trace metals, except molybdenum and uranium, are completely retarded in the mixing zone. This is confirmed by analysis of water samples from the shallow aquifers down- gradient from the tailings which contain only molybdenum and uranium in concentrations above background (White, et al„ 1984).
The calculations by PHREEQE of mixing tailings solution with groundwater predict the precipitation of the calcium molybdate salt, CaMoO4, and uranyl hydroxide, UO2(OH)2. The dissolution of these mineral phases accounts for the molybdenum and uranium in the water extracts. The concentrations of uranium and molybdenum in equilibrium with these minerals, as calculated by PHREEQE, are 100 ppm and 10 ppm, respectively. These are source terms available for succeeding transport.
Regression analysis of the concentrations of trace metals against the iron and aluminum concentrations in acid extracts of soil samples from two locations below the Riverton tailings demonstrates the retardation of trace metals by the precipitation of iron and aluminum hydroxides.
Average concentrations of acid-soluble iron and aluminum in background soils were determined from the concentrations of iron and aluminum in the acid extracts of soils below the tailings. Regression analyses of the trace metals against both iron and aluminum were evaluated, and the regression with the greatest correlation coefficient (R2) is shown in Table 2. The effectiveness of retardation of the trace metals by the precipitation of iron and aluminum was evaluated by comparison of the slope of the regression equations with the ratio of the trace metal to iron or aluminum in the tailings solution (Table 2).
The slope of the regression equations compares well to the ratios in the tailings for the cadmium-aluminum and arsenic-iron couples.
This suggests that the mechanisms operating between the precipitation of aluminum and iron hydroxides and the retardation of cadmium and arsenic remove essentially all of the trace metal from the tailings solution. In the case of nickel-aluminum and uranium-iron, the slopes of the regression lines are less than the ratios in the tailings solutions, which implies that the capacity of the hydroxide precipitates becomes saturated with respect to retardation of nickel and uranium. The small concentrations of nickel in the water-soluble phase below the tailings (Figure 3) reflect this.
The retardation of uranium is more complicated than that of other trace metals. Uranium in the solid phase appears to be partitioned between hydroxide substrates and UO2(OH2)2 as predicted by PHREEQE. The correlation between iron and uranium indicates the retardation by hydroxides. The large positive y-intercept in the uranium-iron regression equation may indicate the presence of a uranium precipitate, independent of iron hydroxide.
The mechanisms of retardation of the trace components by the iron and aluminum hydroxides have not been conclusively
determined. However, the linear relationship between the quantity of major cations (Al+³ and Fe+³) and the trace metals suggests a stoichiometric relationship. It is interesting to note-that the cations (Cd+² and Ni+²) relate most closely to the quantity of aluminum hydroxide; whereas, the anionic species (HAsO4-² and UO2(CO3)2-²) are most closely related to the quantity of iron hydroxide. Substitution of the aluminum ions by the trace metal cations in the crystal structure may account for the retardation of cadmium and nickel. Specific adsorption between the anionic species and the hydroxyl groups of the ferric hydroxide may be the mechanism of anion retardation.