Application of stable isotope tracers to evaluate the hydrodynamic mechanisms of heavy metal mobilization in copper ore tailings

The weathering and erosion of mining and ore processing residues may have a negative impact on local surface waters, which again is likely to be detrimental to groundwater quality. Assessing the environmental response and the dimension of the impacts requires understanding the interaction between the contaminant source and the involved flow systems. The paper presents results of a research on the potential of environmental isotope signatures (δ18O and δ2H) to trace heavy metal emissions from a heap of processing residues via a spring situated at the foot of the heap. The main goal of the investigation was to localize the source of contamination. The heap is considered to be a black box, with no indication given, if the heavy metals in the spring water, with concentrations of up to 2.000 mg/L, are predominantly leached from the processing residues by percolating rain water or by groundwater, entering the heap's base and reappearing at the spring. Knowledge of these processes is essential for the remediation of the site. Thus, the aim of the investigation was to distinguish spring water, which is dominated by local precipitation, from groundwater and to correlate the waters of different origin to the heavy metal contents of the spring water. High contaminant concentrations in the spring water did not allow distinguishing the waters by their chemical composition. In consequence, isotope signatures were used for differentiation. Figure 1 displays the monitored data of spring discharge, rain, δ18O, and δD during a selected observation period in October 2003. In the course of that period, several lighter rain events occurred that did not significantly affect the spring discharge. Also, the isotopic composition of the rain water during the smaller events resembled the average isotopic composition of the spring water so that no significant temporal variation of the isotope data of the spring was observed. In contrast to the minor variations during the lighter rain events, the discharge showed an immediate response to a heavier rain event. At its peak the discharge was about four times higher than before the event. However, the isotopic composition of the spring water indicates that only a part of the additional discharge directly results from percolating rain water. 30 minutes after the first significant increase of the discharge the additional discharge, which is about 100% above normal background values, only contains 30 % of rain water. At its peak, just 10% of the additionally discharged water is rain water. That means that only immediately after the rain event the spring discharge is directly influenced by percolating rain water. During the heavier rain event the lowest metal and sulfate concentrations are reached 30 minutes after the first increase of the discharge, i.e. when the influence of percolating rainwater is most pronounced. The Zn, Cd and SO4 concentrations decrease to 45% of the background concentration, Pb drops even down to 17%. The decrease in concentrations can be explained by dilution. At the peak of the spring discharge, however, the concentrations begin to rise again. That indicates that as soon as the share of percolating rain water in the spring water decreases and the spring water is dominated by discharge from the aquifer, dilution becomes less relevant. This again shows that percolating rainwater only seems to cause a dilution of the contaminants, while the additional discharge of the aquifer triggers additional mobilization of the heavy metals, which occur mainly as sulfates. The results indicate that the source of contamination seems to be located at the base of the heap. While percolating rain water does not mobilize significant amounts of contaminants, groundwater that emerges from the aquifer and migrates at the base of the heap has a strong impact on the heavy metal load of the spring water. Further studies, including those on δ34S signatures along with tracer tests, are planned to verify that idea. If the additional data should confirm the present con cept, it could be concluded, that covering the heap would not be an appropriate remediation measure