In collaboration with McGill University’s Dept. of Mining and Metallurgical Engineering, Noranda Technology Centre (NTC) investigated the prospect of recovering valuable metals from acid mine drainage (AMD) while maintaining effluent quality and reducing the amount of sludge generated. Several chemical methods were evaluated to selectively precipitate and recover metal ions, leading to the development of a conceptual flowsheet for a three-step precipitation process. This flowsheet was further evaluated, along with the associated process economics. The research was funded by Noranda and carried out for the MEND program.

The original three-step precipitation process was developed and investigated in detail at the laboratory scale (Part I) by McGill University. This process consisted of: a) Fe precipitation in the presence of surfactant (dodecylamine, DDA, which was expected to alter the surface properties of Fe(OH)3, thus reducing co-precipitation of Zn); b) sulphide precipitation to obtain a Zn rich sulphide precipitate; and c) a final lime treatment to remove residual metals if, necessary to comply with water quality standards.

Several concerns were raised pertaining to this process. Solid/liquid (S/L) separation and materials handling appear difficult. Furthermore, residual DDA in the effluent may be detrimental to aquatic life. Ultimately, process control and cost are the key problems for sulphide precipitation. As a result, alternative processes were sought.

To this end, three alternatives processes were developed and evaluated by NTC (Part II). In one process, the first step consisted of Fe(III) precipitation with CaCO3, followed by precipitation of metals with NaOH/Na2CO3. In another process, Cu was removed by cementation, using Fe powder. Iron (III) was subsequently precipitated as a phosphate, using H3PO4; and Zn was removed as a hydroxide, using Ca(OH)2. The use of Na2S to obtain ZnS/CuS precipitates was further investigated in a reverse version of the three-step process. Following the removal of Zn and Cu as sulphides at pH 3.5 in the first step, lime neutralization in conjunction with aeration at pH 9.5 was applied to simultaneously precipitate iron and the remaining metal ions, and to produce acceptable effluent quality together in one step.

The results obtained from both studies are:

Part I – McGill Study:
1. With the use of lime, the iron present in AMD can be completely removed as ferric hydroxide at pH 3.5, following oxidation with H2O2. The solid content of the settled sludge ranges from 6 to 8%.
2. The use of DDA in the first step to reduce co-precipitation of other metal ions onto ferric hydroxide sludge slightly improved the subsequent Zn recovery. However, settling of precipitates in all stages deteriorated and the content of the leachable metals in the iron sludge increased. As a result, the use of DDA is not recommended in the process.
3. The iron sludge required several washing cycles in order to remove leachable metals.
4. Zn and Cu can be selectively recovered by using either Na2S, H2S or NaHS in the second step. The Zn/Cu selectivity is closely dependent on pH (e.g. 3.5) and the alkaline reagent (e.g. NaOH) used for pH control. However, as the Zn/Cu recovery increases with increasing pH, the Zn/Cu grade of the sludge decreases.
5. More than 90% Zn recovery and greater than 50% Zn grade can be obtained at pH 4.5 when lime and Na2S are used to set the pH and precipitate the Zn.
6. H2O2, O3 and Trapzene (a CaO2 mixture; patent pending FMC Corp.) were evaluated as Fe oxidants. O3 was technically the most effective when it was used in stoichiometric quantities.
7. Lime treatment of the overflow from the second step to pH 9.5 resulted in an effluent quality similar to that from the conventional lime neutralization process.

Part II – NTC Study:
1. The McGill process was further examined in reverse order and as a two-step process. In this two-step process the removal of Zn/Cu as sulphides at pH 3.5 was performed first followed by the oxidation of iron with air and precipitation with lime at pH 9.5. This method required a large quantity of Na2S (e.g. 3-4x stoichiometric requirement) and technical difficulties in the separation of the Zn-rich sludge were encountered.
2. Of the chemical processes examined, the two-step process yielded the least contaminated iron precipitate and the highest Zn recovery. In addition, this process did not produce sludge requiring special disposal. However, the Zn grade of the precipitate was about 30%.
3. Biological oxidation could oxidize iron, but the required retention time was 2-6 days. On a cost basis, biological oxidation seems to be at least one order of magnitude cheaper than chemical methods.
4. Although the use of CaCO3 resulted in the least contaminated Fe(OH)3 precipitate and its cost was one order of magnitude less than other chemicals tested, the iron still needed to be oxidized before precipitation.
5. Process economics for each process investigated and oxidizing reagent used were assessed. The cost for each process was compared to the cost of the lime neutralization treatment plant being operated at Les Mines Gallen, site of the AMD used in the tests. The most expensive method was the three-step ZnS precipitation, and the least expensive was the two-step CaCO3/NaOH process.
6. Use of PO4(-3)to selectively remove ferric iron, following cementation of copper with iron, and precipitation of zinc with lime was also explored (based on initial Dahnke’s master’s thesis, 1985). The process was not technically or chemically feasible.
7. Requirements for dewatering of each precipitate, generated at each step along the processes were determined. ZnS precipitates required flocculation, clarifier settling and good filtering (e.g. via filter press).

The processes investigated suffer from high costs relative to conventional lime treatment. In particular, the costs of chemical sulphide reagents, and H2O2 are not economical. As a result, further research aimed at optimising the existing flowsheets is not recommended, unless new concepts which radically reduce costs or S/L separation steps are involved.

Some alternative options for developing a process flow sheet should be examined. Suggestions are listed below in order of priority:

I. Following the reduction of all iron in the AMD to ferrous iron with SO2, precipitate ZnCO3 in the first step; then oxidize iron with air and precipitate iron and other residual metal ions with lime.
II. Investigate the biological sulphate reduction process and/or a combined biological/chemical processes.
III. Selective leaching of Zn from lime sludges should also be explored as a potential alternative. Particularly, the sludge generated from the two-step NTC process should be looked at, due to its co-absorption property. Zn selectively leached out from the sludge can be subsequently precipitated with Na2CO3.
Any new process must consider the acceptability of end products from each process for recycling to a Zn roaster, Zn concentrator or lead circuit.