Introduction

Mining and metallurgical activities are characterized by the production of huge volumes of waste rock and finely crushed mill tailing. Production of acid mine drainage can occur by chemical and especially microbiological oxidation of the sulphide phase in piles of waste rock or fine tailings that are in contact with air and water. The resulting water is usually high in acidity and dissolved metals and when the process is started, it’s extremely difficult to stop it. Overall, acid mine drainage is a serious environmental problem with estimated costs of treatment across the planet in the tenths of billions of dollars [1]. On the other hand however, Canada’s extractive metallurgical industry of aluminium produces every year several million tonnes of highly alkaline residues (red mud) that also represent harmful effects to the environment.

Permeable reactive barriers (PRB) are an in situ passive alternative technique to traditional systems like effluent neutralising using limestone. In PRBs, reactive material is placed in the subsurface in the path of a plume of contaminated groundwater by excavating the aquifer material. As the contaminant moves through the reactive material, reactions occur that transform it to less harmful (non-toxic) or immobile species.

This paper presents the results of a research project that is assessing the efficiency of alkaline residues from the aluminium extraction industry (red mud) as reactive material for PRBs.

Experimental setup, materials used

Experimental columns were set up in order to simulate groundwater seepage flow, study the reaction kinetics and ultimately predict the decontamination efficiency and effective lifetime of permeable reactive barriers composed of red mud. The reactive material studied in these experiments is a pelletized form of alkaline residues from the aluminium extractive industry. The pellets were provided by Virotec and are called Bauxsol^TM pellets. The composition of these residues is : 32.3% iron oxides & oxyhydroxides, 18.2% hydrated alumina, 17.5% sodalite, 7.1% quartz, 6.6% cancrinite, 5% TiO,sub>2 , 4.6% Ca(Al) hydroxides & hydroxycarbonates, 3.9% Mg(Al) hydroxides & hydroxycarbonates, 2.3% calcium carbonates (and others 2.5%). They correspond to a metal binding capacity of over 1,500 meq of metals/kg. The pellet form is preferred because of its high porosity, as principal reactive material of the PRB. The CaCO₃ equivalent (alkalinity) of the pellets was estimated to be 0.3 Kg CaCO₃ eq /g [2].

A series of 5 experimental columns in a row was used. Each 100cm long and 7.5 cm internal diameter Plexiglas^TM column (Figure 1) was packed with Bauxsol^TM pellets mixed with silica sand (at various proportions), to ensure a homogeneous matrix and avoid blockage or development of preferential routes during flow of simulated contaminated groundwater. Layers of sand were placed above and below the reactive medium in the column to ensure good flow distribution. Ten sampling ports were installed on each column at distances of 2, 7, 13, 18, 24, 32, 42, 56, 71 and 87 cm. The ports were adjusted to the column wall, and a glass syringe was inserted to allow sampling.

Figure 1: Schematic diagram of an experimental column

In our experiments acid effluents were pumped (using a variable multi-channel speed pump) from bottom to top in order to simulate the average groundwater flow rate in most cases (~1ft/day), as well as higher flow rates in order to account for seasonal variations and predict the effective life time of the barriers. Sampling took place once a week for several months.

Reactive mixes and acid effluents

Due to the low permeability of tailings and red muds, we used only mixes of sand and Bauxsol^TM pellets, under various proportions for the construction of our reactive barrier columns. Table 1 shows the exact composition of each column.

Table 1 : Composition of the reagent mixture of every column

Each column has an internal diameter of 7.5 cm and the material inside the column corresponds to approximately one meter height. A layer of sand was added to both extremities to ensure a good flow distribution.

Mine waste effluents from the Doyon Mine in Val d’Or were used to simulate contaminated groundwater flow. The pH of these effluents is around 2 and contains high concentrations of heavy metals (Table 2). We chose these effluents in order to introduce the “worse case scenario” for the clean up of contaminated plumes in mining sites. It suffices to state here that the Doyon mine spends a few $ million every year for the neutralisation of these effluents and this will continue for at least another 30 to 40 years (i.e. well beyond the end of mine life).

Table 2 : Acid effluent characteristics

Column tests

The effluent is introduced to the bottom of each column by a variable multi-channel speed pump and our goal was to determine necessary retention times to obtain a final pH around 6.5-7.5. We were able to obtain flooding times of our material ranging from 10 min to up to 6h (which corresponds to our minimum pumping rate of 0.5 drips/ sec). At the beginning of our leaching experiments retention times had a wide spread (of up to one hour) but, with time they stabilised and reached an equilibrium.

Columns 1, 4 and 5 seem to be the most efficient. Initially, we observed a strong increase of the pH and subsequently a decreasing period followed by a pH stabilisation. It is probably caused by the presence of highly leachable carbonates and why pH at t=0 min is higher than pH at t=90 min. These results were obtained with a retention time of approximately 1.5h.

Columns 2 and 3 seem less effective and over a period of 24 h the pH obtained was about 4.5. With columns 1, 4, and 5, we could easily maintain the pH around 7 over a long period. However, we noticed a performance improvement of the reagent mixture during the last tests. It may be possible that columns 2 and 3 become more efficient. However, it is still remarkable that column 1 was more efficient with less reactive material. Until now leaching was carried out with a constant flow of effluent.

Contaminant evolution

We observed that the concentration of heavy metals decrease considerably during the reactions in each of the five columns. Figure 2 shows as an example the evolution of Cu, Ni Zn and Pb in column 1 where as Figure 3 shows the Fe evolution in the same column.

The average heavy metal removal efficiencies of Ni²⁺, Cu²⁺, Zn²⁺, Pb²⁺ and Fe²⁺ was 75.5%, 899.%, 79%, 90% and 99.8% respectively. Overall performance of the simulated acid effluent flow through the reactive permeable barrier can be summarised in Table 3. The treated effluents can easily satisfy the current metal mining liquid effluent regulations and guidelines in Canada and Quebec.

Table 3 : Characteristics of the effluent before and after treatment

Figure 2 : Evolution of contaminants in column 1, t = 90 min

Figure 3 : Evolution of Fe in column 1

Conclusions

A new acid remediation technology was presented in this paper whereby metallurgical residues from the aluminium extraction industry can be used to construct permeable reactive barriers (PRB) to treat acid mine effluents. This technology is very promising for treating acid mine effluents and their harmful environmental effects. Based on the laboratory column experiment presented in this paper, it can be concluded that a permeable reactive barrier made of red mud pellets can effectively treat acid mine effluents and easily satisfy the current metal mining liquid effluent regulations and guidelines in Canada and Quebec.

The PRB reduces considerably the pH and heavy metal concentrations to levels well below regulations. Permeable reactive barriers seem to be a very promising technology for the abatement of AMD generation by pyrite tailings or mine waste. They can become a very attractive in situ passive alternative technique to traditional systems like effluent neutralising using limestone. However, further laboratory column experiments should be carried out in order to evaluate their characteristics and effectiveness on a long-term basis.

References

1. Natural Resources Canada, MEND Manual, Report 5.4.2. (2000). CD-ROM.
2. McConchie, D., Clark, M., Davies-McConchie, F., and Fergusson, L., The use of Bauxsol technology to treat acid mine drainage. Mining and Environmental Management, 2002. 10 (4): 12-13.