Utilizing zero-valent iron for the remediation of sulfate-rich mine water

Abstract

Coal and gold mining in South Africa generates sulfate-rich mine water, where Ca and Mg are the dominant cations. Gypsum solubility controls Ca concentration and, to a certain degree, sulfate. Mg solubility is not controlled through precipitation and owing to charge balance principles, if Mg is high, sulfate will be high. Thus, to increase the quality of the mine water S, Ca and Mg must be removed. Various treatment options for the management of mine water exist, however, the passive treatment utilizing zero-valent iron (ZVI) was investigated. The aim was to determine if ZVI can be used to increase mine water quality.

To investigate this, sulfate solutions with concentrations of 300 mg l-1 (3.12 mM) and 1200 mg l-1 (12.5 mM) were prepared as a) CaSO4 solution, b) MgSO4 solution and, c) Ca-MgSO4 solution. These solutions represent typical concentrations of coal mine water chosen for this study. Solutions were further divided into sub-samples (40 ml) and adjusted to a pH of 4 and 8. ZVI (1 g) was added to each sample and dissolved oxygen was purged from the samples using argon gas. Batch sorption experiments were conducted over a period of 145 days, where samples were collected and analysed after days 1, 20, 120 and 145. The concentrations of Fe, Ca, Mg and S were analysed with ICP-AES. Dissolved Fe2+ was measured using the 1,10-phenanthroline colorimetric method. Dissolved Fe3+ was calculated as the difference between total dissolved Fe and dissolved Fe2+. The pH and EC for each sample was measured, as well.

Fisher’s LSD tests showed that S, Ca and Mg in solution, in contact with ZVI, decreased significantly over time. The initially alkaline CaSO4 solution, at 12.5 mM (ALK_CaSO4), showed the greatest decrease of S (19.58 mmol kg-1 ZVI) and Ca (17.87 mmol kg-1 ZVI). Followed by the initially acidic Ca-MgSO4 solution, at 12.5 mM (AC_Ca-MgSO4), for S (15.37 mmol kg-1 ZVI), Ca (5.93 mmol kg-1 ZVI) and Mg (10.68 mmol kg-1 ZVI). Lastly, the initially acidic MgSO4 solution, at 12.5 mM (AC_MgSO4), showed the third greatest decrease in S (14.57 mmol kg-1 ZVI) and Mg (14.76 mmol kg-1 ZVI), where the highest removal of Mg occurred.

Oxidation of Fe0 to surface Fe2+ was enhanced by the increased SO42- concentration. The combination of Ca and Mg in solution, initially acidic, produced the highest concentration of dissolved Fe. Solutions containing Mg also showed lower ratios of Fe3+/FeTot compared to the CaSO4 solutions. Selected XRD analyses showed that in 145 days pyrite, along with goethite and magnetite precipitated on the ZVI’s surface in both the CaSO4 and Ca-MgSO4 solution. For the MgSO4 solution, however, only pyrite was detected.

Practical simulations indicated that removing 20% S, Ca and Mg in 1 m3 of water, in 20 days, from the Ca-MgSO4 solutions required more ZVI, compared to 145 days. ZVI also showed efficacy over the long term. Allowing longer reaction time with simulated mine water (e.g., 145 day) resulted in less ZVI needed per m3 of water.