Mine Drainage Remediation Using Barium Carbonate Dispersed Alkaline Substrate

Abstract

The present invention relates to a treatment system having a barium carbonate (BaCO.sub.3) dispersed alkaline substrate (BDAS) for use in the remediation or at least partial remediation of mine drainage (MD) and/or environmental media contaminated with a source of MD. The invention utilizes chemical, biological and combined treatment systems remove high concentrations of sulfates, hardness, heavy metals and N-compounds, that may exist in the MD as well as high concentrations of alkalinity produced during the remediation process. The invention further extends to a process for treating MD and/or environmental media contaminated with MD and to an apparatus for use in this process.

Claims

1. An apparatus for the bio-remediation, or at least partial bio-remediation, of environmental media contaminated with a source of mine drainage (MD), said apparatus including: (i) a means for introducing environmental media contaminated with a source of MD; (ii) a support matrix; (iii) a barium carbonate BaCO.sub.3 alkaline substrate; (iv) a means for removing sulphate, cation, electrical conductivity (EC) and total dissolved solids (TDS) by precipitation; (v) a bioreaction vessel for containing a microbial consortium; (vi) a means for removing nitrates and cyanide by the bacteria consortium; and (vii) a means for removing the treated environmental media.

2. The apparatus of claim 1, wherein the apparatus is in the form of a fixed-film bioreactor.

3. The apparatus of claim 1, wherein the apparatus is in the form of a fixed-film biocell.

4. The apparatus of claim 2, wherein the apparatus is in the form of an up-flow bioreactor or a down-flow bioreactor.

5. The apparatus of claim 4, wherein the bioreactor is a down-flow reactor with the supernatant open to the atmosphere in order to maximize iron oxidation and to minimize iron (II) mobility in the bioreactor.

6. The apparatus of claim 1, wherein the support matrix serves as an inert physical support mechanism for the microbial community.

7. The apparatus of claim 1 or claim 6, wherein the support matrix also serves as a surface media for the dispersal of the alkaline substrate thereby allowing for the alkaline substrate to take the form of a dispersed alkaline substrate (DAS).

8. The apparatus of claim 7, wherein the dispersed alkaline substrate (DAS) is in the form of a BaCO.sub.3 dispersed alkaline substrate (BDAS).

9. The apparatus of claim 1, wherein the support matrix material may be selected from the group consisting of any one or more of an inert organic medium, wood chips and gravel.

10. The apparatus of claim 1, wherein the bioremediation, or at least partial bioremediation, being utilised is in the form of an active bioremediation system.

11. The apparatus of claim 9, wherein the inert organic material is wood chips.

12. The apparatus of claim 9, wherein the support matrix material consists of a combination of wood chips and quartz gravel.

13. The apparatus of claim 11, wherein the ratio of wood chips to BaCO.sub.3 is 1:4.

14. The apparatus of claim 9, wherein the support matrix provides an electron donor for the bacteria.

15. The apparatus of claim 1, wherein the system being utilised is a passive system.

16. The apparatus of claim 9, wherein the inert organic material is manure.

17. The apparatus of claim 16, wherein the manure is horse manure.

18. The apparatus of claim 1, wherein the microbial consortium consists primarily of sulfate reducing microorganisms which have been introduced into the bioreaction vessel as a sulfate reducing microorganism inoculum.

19. The apparatus of claim 18, wherein the inoculum is an enrichment culture prepared from a solid sample of sewage sludge.

20. The apparatus of claim 1, wherein the barium carbonate BaCO.sub.3 contributes to the acidity reduction and pH stabilization of the environmental media being treated.

21. The apparatus of claim 1, wherein the barium carbonate BaCO.sub.3 assists with the immobilization of sulfides in the bioreactor.

22. The apparatus of claim 1, wherein the barium carbonate BaCO.sub.3 has a pH of about pH 3 to about pH 8.5.

23. The apparatus of claim 1, wherein a zero valent iron (ZVI) is included.

24. The apparatus of claim 23, wherein the ZVI serves as an energy source for the sulfate reducing bacteria.

25. The apparatus of claim 23, wherein the ZVI interacts chemically with the H.sub.2S generated by the sulfate reduction.

26. The apparatus of claim 23, wherein the ZVI is involved in the direct reduction of the metals present in the environmental media being treated.

27. The apparatus of claim 9 and claim 23, wherein the ratio of inert organic material to BaCO.sub.3 to Zero Valent Iron (ZVI) to sulfate reducing microorganism inoculums is 60%:19%:1%:20%.

28. The apparatus of claim 2, wherein the fixed-film bioreactor operates under aerobic conditions.

29. The apparatus of claim 2, wherein the fixed-film bioreactor operates under anaerobic conditions.

30. The apparatus of claim 1, wherein the bioreactor operates with an initial flow rate of 1.09 mL/minute and a hydraulic retention time of about 24 to about 48 hours and wherein the flow rate can be increased up to 2.5 L/min and the retention time can be decreased up to 9 hours.

31. The apparatus of claim 1, wherein the bioreactor operates at an oxidation reduction potential (Eh) of between about −200 mV and about −250 mV.

32. The apparatus of claim 1, wherein the bioreactor can operate at either continuous or pulsed flow.

33. The apparatus of claim 32, wherein the operation occurs under pulsed flow.

34. The apparatus of claim 1, wherein the precipitation is in the form of any one or more of Barium sulfate (BaSO.sub.4), carbonates ((Me.sup.2+)CO.sub.3) and oxy-hydroxides (Me.sup.2+/3+O(OH)).

35. A process for the bioremediation, or at least partial bioremediation, of environmental media contaminated with MD, wherein the process includes the step of removing environmental media from a mine drainage contaminated site and exposing the environmental media to a mixture including an indigenous microbial consortium of microorganisms, as identified herein, a wood chips as a minimal source of electron donor, as identified herein and a dispersed alkaline substrate, more particularly dispersed barium carbonate BaCO.sub.3, for a sufficient period of time so as to allow for the chemically and biologically mediated precipitation of the sulfate, nitrate, phosphate and metals.

36. A treatment process for treating acid mine drainage (MD) contaminated environmental media, said process including the following stages: The first stage including the steps of: (i) providing a bioreactor wherein the bioreactor includes a wetted support matrix which has been pre-treated with BaCO.sub.3 to form a BaCO.sub.3 dispersed alkaline substrate; (ii) introducing the MD contaminated media into the bioreactor and subsequently introducing the indigenous microorganisms; and (iii) controlling the hydraulic retention time of the bioreactor such that the hydraulic retention time is between about 9 hours and about 24 hours. The second stage including the steps of: (iv) allowing for the precipitation of the sulfate present in the source of the MD contaminated environmental media as barium sulfate (BaSO.sub.4); (v) allowing for the precipitation of Ca, Mg, Na and trace metals present in the source of the MD contaminated environmental media as calcite and aragonite (CaCO.sub.3); (vi) allowing for the precipitation of heavy metals present in the source of the MD contaminated environmental media as oxy-hydroxides (Me.sup.2+/3+O(OH)); and (vii) removing cyanide and nitrates within the biological processes involved in the invention.

37. Barium carbonate (BaCO.sub.3) for use in a process for the bioremediation, or at least partial bioremediation, of environmental media contaminated with MD in combination with bacterial treatments.

38. The apparatus of claim 1, substantially as herein described with reference to the accompanying figures.

39. The process of claim 35, substantially as herein described with reference to the accompanying figures.

40. The treatment process of claim 36, substantially as herein described with reference to the accompanying figures.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0082] FIG. 1: depicts the temporal evolution of the physicochemical parameters from 0 to 168 h and its relation with drinking water standards (SANS 241);

[0083] FIG. 2: are graphs depicting temporal evolution of cations and anions from 0 to 168 h;

[0084] FIG. 3: shows difractograms of E1 and K1 at 0, 6 and 168 h. W: witherite, B: barite, C: calcite and A: aragonite;

[0085] FIG. 4: shows the schema and description of the DAS-BaCO3 bioreactor;

[0086] FIG. 5: is a graph depicting physicochemical parameters (pH, sulfates, Sal, TDS, Cond and Eh) of the samples collected from the outlet of the three columns (A, B and C) over time;

[0087] FIGS. 6a to 6c: is a graph depicting the spatial evolution of the main chemical compounds of the water throughout column C, from top (inlet) to bottom (port);

[0088] FIG. 7: is a graph depicting XRD difractograms of the three samples;

[0089] FIG. 8a: shows the dissolved oxygen in a BaCO3 experiment;

[0090] FIG. 8b: shows the nitrate reduction in the BaCO3 experiment;

[0091] FIG. 9: is a graph showing cyanide removal from a treated sample after 120 hours of flow through SRB-DAS-activated charcoal cartridge-BDAS;

[0092] FIG. 10: shows a cascade for aerating the system;

[0093] FIG. 11: shows the integration of the PLC system with the whole pilot plant;

[0094] FIG. 12a: is a graph depicting the variation of pH over time as obtained from chemical analysis of the treated media;

[0095] FIGS. 12b and 12c: is a graph depicting data from chemical analysis of the treated media;

[0096] FIG. 12d: is a graph depicting sulfate concentration over time as obtained from chemical analysis of the treated media;

[0097] FIGS. 12e to 12h: is a graph depicting data from chemical analysis of the treated media;

[0098] FIG. 13: SEM-EDS images of reactive materials (whiterite);

[0099] FIGS. 14a and 14b: is a graph depicting data obtained from the PLC of tanks 1, 4 and outlet, after replacement of reactive material chemical analysis of the treated media;

[0100] FIGS. 15a to 15e: is a graph depicting data as obtained from chemical analysis of tanks 1, 4 and outlet after replacement of reactive material; and

[0101] FIGS. 16a to 16f: is a graph depicting the evolution of the physicochemical parameters (on average) of the contaminated media throughout the pilot plant, from inlet (JD) to outlet (AS).

[0102] The presently disclosed subject matter will now be described more fully hereinafter with reference to the accompanying Examples, in which representative embodiments are shown. The presently disclosed subject matter can, however, be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the embodiments to those skilled in the art.

EXAMPLES OF THE INVENTION

[0103] The invention was performed in accordance with the following steps: [0104] Phase 1: Batch experiments were performed with acid and non-acid mine drainage to showcase the chemical interaction between MDs and barium carbonate. [0105] Phase 2: Lab scale bioreactors were constructed in the form of down flow column experiment filled with barium carbonate dispersed alkaline substrate to treat acid and non-acid mine drainage. [0106] Phase 3: Pilot scale of water treatment plant were designed and installed at two different sites, where acid and non-acid mine drainages were treated.

Phase 1: Batch Experiments

1.1 Introduction:

[0107] The Acid Mine Drainage (AMD) generated from pyrite's oxidative dissolution, typically contains high concentration of anions (SO.sub.4.sup.2−) and metal (mostly Fe.sup.3+>Al.sup.3+>Cu.sup.2+>Zn.sup.2+>Mn.sup.2+) which makes it a significant environmental problem for South Africa, as well as for other mining countries.

[0108] The South African mine drainages (MD) is characterized by a wide pH range from acidic (2.6) to alkaline (8). The main reason for this fact is that the host rock contains mainly pyrite and carbonates (such as dolomite). Therefore the MD is characterized by having high salinity (Ca>Mg>Na), hardness and heavy metal concentrations such as Fe.sup.3+>Al.sup.3+>Mn.sup.2+ and moderate to low trace metal concentrations such as Ni.sup.2+>Zn.sup.2+>Cu.sup.2+.

[0109] Therefore, the conventional passive chemical systems based on a CaCO.sub.3 or MgO neutralization process are not completely effective for these leachates, because: (1) the acid mine drainage treatment by CaCO.sub.3 or MgO allows the neutralization and removal of heavy metals. However, it increases the salinity and hardness in the treated effluent. (2) The low solubility of CaCO.sub.3 at high pH limits its use in treating acid and not alkaline drainages. Also the active systems, such as reverse osmosis or GYP-CIX, can remove salinity and hardness. However, the high maintenance costs and the brine generated by the treatment decreases the viability of these systems.

[0110] Based on hydrogeochemical characteristics of this type of leachate, many treatment systems have been showcased that are generally based on sulfate-reduction bioreactors. This technology, despite having been optimized in recent years, has not been able to completely remove the high concentration of SO.sub.4.sup.2− and it did not decrease salinity and hardness in these leachates. BaCO.sub.3 was tested in simple batch experiments in the 1970's due to its dissolution in a wider range of pH (0-9) and due to its capability to precipitate sulfate as BaSO.sub.4, but it was not considered viable because the dissolution rate was very low at pH values of 7-10. In the 80's, 90's and again in 2006, BaCO.sub.3 was tested as a step in an active process to remove sulphate. However, these studies did not optimize the BaCO.sub.3 concentration, residence time nor provided relevant information about the geochemical behaviour of this compound and its use in MD treatment.

[0111] Current studies have shown that BaCO.sub.3 has a good dissolution rate between pH values of 0-6.5 and that the dissolution rate decreases when pH increases. In addition, it was also shown that BaCO.sub.3's dissolution rate increases with increasing temperature because of its endothermic nature. Moreover, previous studies showed variations between theoretical thermodynamics and experimental results regarding the dissolution of the BaCO.sub.3. This knowledge is extended in this research which focused on addressing these issues by conducting a geochemical study with BaCO.sub.3 and MD that could explain both its behaviour as well as its potential to remediate these leachates. Understanding these processes will allow the optimization of BaCO.sub.3 usage for sulfate removal and its contribution in removing salinity and hardness from acid and alkaline MD.

1.2 Experimental materials

[0112] Two drainages with different hydrogeochemical characteristics from active and abandoned mines were collected from the South African provinces of Mpumalanga (25°42′20.4″S 29°59′28.4″E) and Gauteng (25°50′10.0″S 29°14′03.7″E) which were used as natural reagent solutions for batch experiments. The first drainage was an alkaline mine drainage (AMDE), whose hydrogeochemical characteristics conforms to the average of typical coal mine drainages (high sulphate, salinity and hardness concentration). The second drainage was acid mine drainage collected from an abandoned mine (AMDK), which is characterized by high acidity and pollutant concentration. Each sample was taken on site in polyethylene tanks (ca. 260 L) for further experiments and part of each sample (1 L/AMD) filtered through a 0.45 μm filter within 24 h for chemical analysis.

[0113] Alkaline material used in this experiment was BaCO.sub.3 (Protea Chemicals Company SA). BaCO.sub.3 have a purity of 88.6%. These materials contain impurities including Fe and S as SO.sub.4.sup.2−, in negligible concentrations.

1.3 Batch Experiment

[0114] Batch experiments were conducted to test the interaction of alkaline material with AMDE and AMDK at different time intervals (0 min, 5 min, 15 min, 40 min, 2 h, 6 h, 12 h, 24 h, 36 h, 48 h, 72 h, 96 h, 120 h, 144 h and 168 h) in falcon tubes (50 mL) under continuous mixing in a rotary mixer at 12 rpm and room temperature. Four series of interactions were carried out using solid:liquid (w/w) ratios of 1:400, 1:57 and 1:160, 1:80 for experiments with AMDE and AMDK, respectively. Each interactions will be identified throughout the paper as E1 that refers to the interaction between 40 mL of AMDE and 0.1 g BaCO.sub.3; E2 to the interaction between 40 mL of AMDE and 0.7 g of BaCO.sub.3; K1 to the interaction between 40 mL of AMDK and 0.25 g of BaCO.sub.3 and K2 to the interaction between 40 mL of AMDK and 0.5 of BaCO.sub.3. At the end of each time interval, the tubes were removed from the rotary mixer and the supernatant was separated from the solid product by centrifugation at 4000 rpm for 3 min. Finally, the supernatant solutions were filtered through a 0.45 μm filter and the solid product was dried at 40° C.

1.4 Chemical Analysis

[0115] The following parameters were analysed on site from the collected samples to avoid the dissolution effects of the CO.sub.2 (g) and O.sub.2 (g): pH, Electrical Conductivity (EC), salinity (Sal), redox potential (Eh) and temperature (T). The pH, EC, Sal and T were measured with the ExStix®II multi-probe, while Eh was with ExStix®II ORP (Pt and Ag/AgCl electrodes) probe. The Eh measurements were then corrected to standard hydrogen electrode (SHE). Samples were filtered and acidified to pH<2 with HNO.sub.3 (2%) and stored at 4° C. for further chemical analysis at the Institute for Ground Water Studies, University of the Free State. Sulfate concentrations were analysed by a portable Hach spectrophotometer (model DR/900 colorimeter) according to the turbidimetric method described in the Hach Procedures Manual-Method Sulfate 608. Fe.sup.2+ and Fe.sub.Total were determined after filtration (0.45 μm) with a Hach spectrophotometer (model DR/900 colorimeter) according to the colorimetric method described in the Hach Procedures Manual-Method Ferrous iron 255 and FerroVer 265. All these chemical analysis also were carried out on site.

[0116] The neutralization potential of BaCO.sub.3 was determined by treating a sample with a known excess of standardized hydrochloric acid subjected to heat treatment (95° C.). Finally, the amount of neutralizing bases expressed in tons CaCO.sub.3 equivalent/thousand tons of material was determined from the amount unconsumed acid by titration with standardized sodium hydroxide.

[0117] The BaCO.sub.3 was digested by an aqua regia solution (1HCl:1HNO.sub.3:1H.sub.2O) at 90° C. for 1 h up to its complete dissolution. Total Element Concentration (TEC) from the digestion, as well as the sub-samples, were analysed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES; Jarrel Ash Atom comp 975). The mineralogical characterization of the final experimental products was carried out by X-ray diffraction (XRD, powder method) using a Panalytical Empyrean diffractometer under following conditions: slit fixed at 10 mm, Cu/Kα monochromatic radiation, 40 mA and 45 kV. Samples were run at a speed of 2°θ/min (5-70°). The spectrum was obtained by Highscore software. In addition, solid samples were also studied using a scanning electron microscope equipped with an energy dispersive system (SEM-EDS; JEOL model GSM 6610).

1.5 Geochemical Modelling

[0118] Precipitation of newly formed solid phases by the BaCO.sub.3 dissolution could control the fate of the metal concentrations in both the acid and alkaline mine drainage, studied by the batch experiment. The results of the hydrogeochemical analysis from supernatant of each reaction (sub-sample) were modelled by PHREEQC-2 geochemical speciation model using MINTEQ thermodynamic database to predict the aqueous speciation of leachates and saturation indices of solid phases in the experiments [SI=log(IAP/KS) where IAP is the ion activity product and KS is the solubility constant]. Zero, negative or positive SI values indicate that the solutions are saturated, undersaturated and supersaturated, respectively, with respect to a solid phase.

1.6 Results

[0119] Results of hydrogeochemical characterization of the AMDs are reported in Table 1. The main difference between the two mine water samples is the pH. The pH values of AMDE and AMDK were 8.2 and 2.93, respectively. In the case of AMDK, low pH values were related to the low carbonate concentration in the host rock, which contain high sulphide concentration. Its intense oxidation and subsequent dissolution of pyrite, produces a large amount of acidity. In the case of the AMDE it had circum-neutral to alkaline pH-values due the low content of sulphide minerals and the presence of carbonate or basic silicate minerals. The carbonate dissolution also contributes to lowering the water quality by increasing the hardness and salinity, which also affects the ecosystem.

TABLE-US-00001 TABLE 1 Significant physicochemical parameters of the acid and alkaline mine drainages; AMDE AMDK pH 8.2 2.93 EC (mS m.sup.−1) 209 170 Redox potential (mV) 295 415 Ca (mg/L) 256.0 169.84 Mg (mg/L) 138.9 66.34 Na (mg/L) 12.18 41.30 Ba (mg/L) 0.040 0.028 Fe (mg/L) 0.042 34.24 Al (mg/L) 0.019 44.89 Sulfate (mg/L) 1250.0 1400 Mn (mg/L) 0.023 10.11 Zn (mg/L) 0.016 1.31

[0120] The neutralization potential of BaCO.sub.3 obtained was 525 tons CaCO.sub.3 equivalent/thousand tons of material. The neutralization potential of BaCO.sub.3 is lower than calcite which has a high neutralizing capacity of 937.5 tons CaCO.sub.3 equivalent/thousand tons of materials. However, the calcite is scarcely soluble at circum-neutral pH (6-7), while BaCO.sub.3, despite having a low solubility at circum-neutral pH (6-7) is able to dissolve at pH values of up of 8-9. Total Element Concentration (TEC) confirmed the product data from Protea Chemicals, which indicated that the most significant impurities were S and Fe with values of 0.30% (total sulfur as SO.sub.4.sup.2−) and 0.004% (Fe total). The average particle size was 1-3 μm.

[0121] Preliminary batch experiments were carried out to test the dissolution capacity of the BaCO.sub.3 in alkaline and acid mine drainage. The results obtained in these experiments showed a sulfate removal percentage of 90% on average and an increase to pH of 9. The BaCO.sub.3 had a higher dissolution at lower pH such as 4-5, whereas, at higher pH such as 8.9 the dissolution of BaCO.sub.3 was slower. However, the dissolution of BaCO.sub.3 after 24 h showed the same behaviour in both AMDK and AMDE, indicating that the pH does not decrease the dissolution of BaCO.sub.3 after 24 h.

[0122] The hydrogeochemical evolution as a function of time of the physicochemical parameters such as pH, Eh, EC, Sal, as well as sulfate concentration in the four ratio (w/v) interactions are shown in FIG. 2. The neutralization potential of BaCO.sub.3 allowed the pH to increase from 2.93 to 8.27 for the K1 and K2 interactions (0.25 and 0.5 g of BaCO.sub.3), and from pH 8.2 to 9.98 on average for the E1 and E2 interactions (0.1 and 0.7 g of BaCO.sub.3), respectively. The Eh values decreased from 295 to 67 mV and from 415 to 128 mV on averages, whereas EC decreased to 942 μS/cm and 1091 μS/cm (variation ±5%), for the E1-E2 and K1-K2 interactions, respectively,

[0123] The decrease in EC values reflects an improvement in the quality of MDs that was confirmed by the decrease in sulfate concentration in the solution. In the experiments with AMDK all these parameters achieved a steady state in 6 h in both interactions (K1 and K2). The behaviour of BaCO.sub.3 was different for the interactions with AMDE (E1 and E2), where a steady state was achieved after 24 h. The sulfate concentrations decreased slowly after 24 h (E 1 and E2 reached 280 and 120 ppm after of 168 h) without achieving a steady state, while in the K1 and 2 interactions, the sulfate concentration was completely removed after 24 h. BaCO.sub.3 dissolution was faster in the K2 interaction where the pH increased from pH 2.93 to 6.79 and the interaction was almost immediate. However the Sal and EC evolution was slower.

[0124] The evolution of metals and sulfates are closely related to the dissolution rate and the concentration of BaCO.sub.3(FIG. 2). Therefore, optimization and understanding of its behaviours is vital to assess its remediation potential. The interaction with a concentration of BaCO.sub.3 larger than 0.1 g (E1) showed higher concentration of dissolved Ba.sup.2+ at the end of the experiment (0.33, 6.7, 4.1 mg/L in E2, K1 and K2, respectively) which did not react during the experiment. Therefore, the concentration of BaCO.sub.3 used in E1 can be considered as the optimum to be used in passive and active systems with a residence time of 24 hours, at most, to get an 86% sulfate removal rate.

[0125] The hydrogeochemical behaviour of the cations, such as Ca.sup.2+, Mg.sup.2′ and Na.sup.+ over time was similar, between E1 and E2, as well as between K1 and K2. Ca decreased drastically within 6 h, the removal reached 97% in the E interactions, but in K interactions took 120 h to reach 51% of Ca.sup.2+ removal. The concentration of Na.sup.+ only decreased 18% in K interactions.

[0126] The evolution of metals during the experiment will only be described and discussed with regards to the K interactions, due to the insignificant concentration of metals in AMDE. The concentration of metals in AMDK was as follow, Al.sup.3+>Fe.sup.3+>Mn.sup.2+>Zn.sup.2+ (44.89>34.24>10.1>1.3, respectively). The removal of Fe.sup.3+, Al.sup.3+ and Zn.sup.2+ were 100%. However the removal of Mn was 66% in 24 h and 86% in 120 h.

[0127] Parameters such as EC, Sal and hardness decreased in all the interactions to values below the allowable limits for drinking water (South African National Standard (SANS) 241, 2006; 2011) (FIG. 1). The removal of SO.sub.4.sup.2-, Ca.sup.24 and heavy metals was the main reason for those parameters to decrease. Most of the passive systems are not able to remove Ca.sup.2+, but increase its concentration (such as the systems based on CaCO.sub.3), however this system has demonstrated its effectiveness in removing anions (SO.sub.4.sup.2−) and cations (mainly Fe.sup.3+, Al.sup.3+, Mn.sup.2+, Zn.sup.2+ and Ca.sup.2+) which is also reflected in the concentration of Sal and EC of the drainage.

[0128] The precipitates collected at the end of the experiment from E1 and K1 interactions, were analysed by XRD (FIG. 3). The analyses showed mainly mineral phases related to the dissolution of BaCO.sub.3 as well as to the precipitation of sulfate and Ca.sup.2+. The geochemical processes involved in the increased of pH, as well as the sulfate, Ca.sup.2+, Mg.sup.2+ removal, including Fe.sup.3+, Al.sup.3+, Mn.sup.2+ and Zn.sup.2+ has been represented by the following equations:

[0129] 1. Representation of dissolution of BaCO.sub.3 in AMD:

BaCO.sub.3(s).fwdarw.Ba.sup.2+(aq)+CO.sub.3.sup.2−(aq)  (15)

[0130] 2. pH values were increased by releasing OH.sup.− radicals and formation of CO.sub.2 that could act as a buffer to control the increase of pH.

CO.sub.3.sup.2−(aq)+H.sub.2O(l).fwdarw.HCO.sub.3.sup.−(aq)+OH.sup.−(aq)  (16)

HCO.sub.3.sup.2−(aq)+H.sub.2O(l).fwdarw.H.sub.2CO.sub.3.sup.−(aq)+OH.sup.−(aq)  (17)

[0131] 3. The increased pH values would allow the trivalent and divalent metals precipitation as oxy-hydroxides and/or oxy-hydroxysulfate of Fe.sup.3+ and Al.sup.3+ and carbonates of Mn.sup.2+ of Zn.sup.2+, respectively. In addition, the presence of carbonates and bicarbonates in solution would promote the Ca.sup.2+ and Mg.sup.2+ removal as carbonates and thus reduce the hardness of these AMDs.

Me+HCO.sub.3.sup.−MeCO.sub.3+H.sup.+  (19)

Me+H.sub.2CO.sub.3.sup.−.fwdarw.MeCO.sub.3+2H.sup.+  (20)

[0132] 4. While the sulphate precipitate like barite (BaSO.sub.4)

Ba.sup.++SO.sub.4.sup.2−.fwdarw.BaSO.sub.4  (21)

[0133] The estimated percentage of those mineral phases were, according to the contact time (0 h, 6 h and 168 h), as follow: E1: 0 h: witherite (71.2%)>calcite (15.9%)>barite (12.9%); 6 h: barite (63.8%)>witherite (26.5%)>calcite (9.7%); 168 h: barite (65.7%)>calcite (19.2%)>witherite (16.9%). K1: 0 h: witherite (76.2%)>barite (13.5%)>calcite (10.3%); 6 h: witherite (71.4%)>barite (18.9%)>calcite (9.6%); 168 h: witherite (53%)>barite (28.5%)>calcite (18.5%).

[0134] However these mineral phases could be masking other sub-idiomorphic or amorphous crystals, mainly in the K interactions, where the metal concentrations were high. This was corroborated by SEM-EDS analyses, where Fe.sup.3+, Al.sup.3+ and Mn.sup.2+ were detected in the precipitates. The thermodynamic simulation with PHREEQC also supported this hypothesis by predicting the precipitation of Fe.sup.3+ and Al.sup.3+ as oxy-hydroxysulfate, poorly crystallized according to XRD analyses. This acted as a sink for trace elements and contributed to reaching the requirements for drinking water. The minerals phases of Mn.sup.2+ and Zn.sup.2+ were not predicted to be saturated by PHREEQC, however both metals were 100% removed from the AMDs. This again demonstrated that there are several discrepancies between the theoretical thermodynamic fundaments and the real geochemical data acquired throughout the experiment. Finally, the improvement of the quality of the MDs used in the four interactions has been so effective that the final concentration of the sulfates was within the limit allowable for drinking water (South African National Standard 241, 2006; 2011).

1.7 Conclusion

[0135] Batch experiments were conducted with the aim to study the behaviour and optimize the use of BaCO.sub.3 in MD remediation. Four interactions were carried out with two different MDs and four different ratios (w/w) BaCO.sub.3: MD (1:400 and 1:57 with AMDE (alkaline) and 1:160 and 1:80 with AMDK (acid)). Each interaction was composed of 15 sub-samples, each of them with different contact time between MD and BaCO.sub.3 (from 0 to 168 h). All the samples achieved a steady state between 6 and 24 h. However the low solubility of the BaCO.sub.3 at high pH slowed down the dissolution in E interactions, where the pH reached up 9.98 and the dissolution continued after 168 h. Nevertheless, E1 interaction reached a sulfates removal of 86% between 6 and 24 h. The sulfates and Ca removal were the most meaningful results in E interactions. Moreover, the total metal removal in K interactions was the determining factor for the improvement of the water quality. According to these results, the ratio used in the E1 interaction can be considered as the optimum to be used in systems with a residence time of 24 hours.

[0136] XRD and SEM-EDS analyses corroborated the sulfates and metals evolution over time by the identification of crystalline and amorphous mineral phases. The modelling also predicted the precipitation of mineral phases such as barite, calcite and Fe/AI oxy-hydroxides. However there were discrepancies between the predictions and the data acquired from the experiments, such as the removal of Zn and Mn that probably were precipitated as carbonates. Therefore the BaCO.sub.3 dissolution varies according to the pH and the composition of the MD. However, at the end of each experiment the water was within the South African National Standard for drinking water.

Phase 2: Column Experiments

2.1 Introduction:

[0137] South Africa has 95% of Africa's known coal reserves and the 9.sup.th biggest recoverable coal reserves (61000 Mt) in the world where 27400 Mt were proven coal reserves in 2012.

[0138] These coal deposits have about 4% of pyrite which is the cause for the coal mine drainage to contain sulphur. However the typical acidity produced by the oxidation of pyrite (equations 22 and 23) and by the subsequent oxidation and precipitation of Fe (equation 24 to 26) is neutralized by the CO.sub.3.sup.−2 released from the calcite (CaCO.sub.3) and dolomite (CaMg(CO.sub.3).sub.2) that is contained in South African's coal; about 6.7% and 10.1% respectively. Therefore, coal mines in South Africa can generate acid, neutral or alkaline mine drainage (MD). When pyrite and other sulfide minerals associated with coal deposits are exposed to water and oxygen, several chemical and biochemical reactions take place.

[0139] Oxidation of pyrite can be produced by oxygen (equation 22) or ferric iron (equation 23) in the presence of water. Further oxidation of Fe.sup.2+ to Fe.sup.3+ occurs when sufficient oxygen is dissolved in the water or when water is exposed to atmospheric oxygen (equation 24). This reaction is also accelerated by the presence of oxidizing bacteria such as Acidithiobacillus ferrooxidans. Ferric iron can either precipitate as Fe(OH).sub.3, (equation 25) or it can react directly with pyrite to produce more ferrous iron and acidity as shown by equation 26. The presence of alkaline compounds such as calcite and dolomite decreases the acidity of the MD by consuming protons (H.sup.+) and releasing bicarbonate anion (HCO.sub.3.sup.−) as shows the equation 27 and 28.

##STR00001##

[0140] Therefore the alkalinity of the coal mine drainage depends, among others, on the ratio between acidic and alkaline minerals of each specific coal deposit and surroundings.

2.2 Case Study

[0141] The case study was done on the alkaline drainage generated by a coal mine situated at (25°42′23.2″S, 29°59′32.7″E), that is mining the coal from the north eastern coalfield of the Karoo basin, located outside in Mpumalanga. The MD generated is collected in the evaporation dam located SW within the facility area (25°42′20.4″S 29°59′28.4″E). This MD has a pH of 7.45 and, in contrast, the electrical conductivity (EC), salinity (Sal) and total dissolved solids (TDS) are fairly high (2090 .sup.μS/cm, 980 mg/L and 100 mg/L, respectively). The MD has high concentrations of sulfates and nitrates (1 253 mg/L and 3 032 mg/L respectively) as well as dissolved Ca and Mg (262.41 mg/L and 132.60 mg/L respectively).

2.3 State of the Art MD Treatments

[0142] Many passive and semi-passive treatments have been developed over the past three decades to remediate MD, such as aerobic and anaerobic wetlands, Anoxic Limestone Drains (ALD), limestone sands, beds, ponds and open channels, diversion wells, reducing and alkalinity producing systems (RAPS), ReRAPS, water-powered devices, windmills, sodium carbonate briquettes, sodium hydroxide, hydrated lime and quick lime. The reactor system that the authors have developed on a laboratory scale is based on a modified Dispersed Alkaline Substrate (DAS) system. The modification includes the substitution of limestone (CaCO.sub.3) with barium carbonate (BaCO.sub.3) powder. This system, called B-DAS, has been designed with the aim to improve the removal of sulfates by precipitating it as BaSO.sub.4 as well as improving the salinity (see reactions below). The aim is extended to find a system that is able to remediate not only acid mine drainage but neutral and alkaline mine drainages as well. BaCO.sub.3 easily dissolves at a pH above 4 which makes it ideal to treat these drainages. The MD used in this study has a pH of 7.45 which undergoes the dissolution process as follow; BaCO.sub.3 is dissolved (equation 29). Dissolved sulfates can precipitate as barium sulfates (equation 30). The pH is increased to 10 by consuming protons and releasing hydroxide anion (equation 29) and bicarbonate anion (equation 31). The high pH and the presence of bicarbonate anions promote metal precipitation as carbonates (equation 32) (e.g. Ca and Mg):

BaCO.sub.3+H.sub.2O.fwdarw.Ba+HCO.sub.3.sup.−+OH.sup.−  (29)

BaCO.sub.3+H.sub.2SO.sub.4.fwdarw.BaSO.sub.4+H.sub.2CO.sub.3.sup.−  (30)

BaCO.sub.3+H.sup.+.fwdarw.Ba+HCO.sub.3.sup.−  (31)

Me+HCO.sub.3.sup.−.fwdarw.MeCO.sub.3+H.sup.+  (32)

2.4 Column Experiment

[0143] Three down-flow columns were constructed from PVC pipes (10 cm inner diameter, height 50 cm) and equipped with four additional lateral sampling ports. Each port had a small perforated pipe in the column matrix to promote homogeneous samplings and allow homogeneous flow within the columns by increasing the area of sampling while, avoiding, as far as possible, preferential flow.

[0144] Each column contained a layer of quartz gravel (particle size about 5-8 mm) at the bottom (2.5 cm). This layer was covered with a 40 cm reactive material layer, which consisted of BaCO.sub.3 and wood shaving mixture. Each column had different ratios of wood: BaCO.sub.3 (w/w); these were columns (A) 1:2 (260 g:520 g) (B) 1:3 (240 g:720 g) and (C) 1:4 (220 g:960 g).

[0145] During the six months of the experiment the down-flow bioreactors, with supernatant open to the atmosphere, were fed with the MD, as input water from the top using a peristaltic pump and flowed down gravitationally. The outflow was collected in a container that also functioned as an aeration and sedimentation tank. The flow rate was 1.09 mL/min with a residence time of 24 hours for each B-DAS columns. The porosity of the systems was 70% (volumetrically calculated).

2.5 Indigenous Bacteria Communities

[0146] Several indigenous communities of microorganisms are always present in MDs. These microorganisms are settled into the bioreactors thanks to the conditions promoted by the barium carbonate dispersed alkaline substrate. The dissociation of barium carbonate decrease the oxidation reduction potential and increase the pH, resulting in the favourable condition for the bacterial settling. The woodchips disposed into the bioreactors is used by the bacteria communities as minimal electron donor, while the acceptor donor is the cations and anions dissolved within the MD.

2.6 Wood Chips and Gravel

[0147] Both materials were used as the inert material within the column. Both materials provide porosity to the system. The wood chips are the organic matter in the system and it represents a minimal carbon source and support for the bacteria settlement. However, the resin acids, a group of diterpenoid carboxylic acids present mainly in softwoods were reported to be toxic to microorganisms.

2.7 Sampling

[0148] During the first week of the columns running, samples were taken daily from the outlet of each column (A, B and C) and after that sampling was done weekly for the next six. Another set of samples were taken monthly from the four sampling ports of each column, to evaluate the spatial evolution of each column during the experiment.

[0149] Finally, the columns were drained and column C was cut with an angle grinder to have access to the precipitates formed on the wood shavings. Three samples of precipitates were collected at the top, middle and bottom of the column for further analysis.

2.8 Analytical Techniques

[0150] Source water was collected from the evaporation dam in 25 L carboys, transported to the laboratory and stored at 4° C. pH, EC, Sal, TDS, redox potential (ORP) and temperature (T) was measured on site. These physicochemical parameters were also analysed from the columns weekly and monthly. The measures were done with the ExStix®II multi-probe and ExStix®II ORP probe. ORP measures were corrected to the Eh standard hydrogen electrode (SHE). Samples were analysed by ICP at the Institute for Groundwater Studies at UFS, filtered and acidified to pH<2 with HNO.sub.3 2% (v/v), to compare influent and effluent chemistry of the columns. Sulfate (SO.sub.4.sup.2−), Fe.sup.2+ and Fe.sup.Total concentrations were analysed by a HACH spectrophotometer (model DR/900 colorimeter) according to the colorimetric methods described in the HACH Procedures Manual (Method Sulfate 608, Method Ferrous iron 255 and FerroVer 265, respectively).

[0151] The precipitation of newly formed solid phases by the BaCO.sub.3 was confirmed by using a thermodynamic model (PHREEQC) as well as by characterizing the final solid products. These saturated mineral phases in the system were estimated, assuming that the initial solution in contact with an alkaline material (in our case BaCO.sub.3) reaches equilibrium with that material. The PHREEQC-2 geochemical speciation model (Parkhurst & Appelo, 2005) in conjunction with the MINTEQ thermodynamic database (Allison et al., 1991) was used to determine the aqueous speciation of solutions and saturation indices (SI) of solid phases that could control the concentration of dissolved species in the simulation SI=log(IAP/KS) where IAP is the ion activity product and KS is the solubility constant. Zero, negative or positive SI values indicate that the solutions are saturated, undersaturated and supersaturated respectively, with regards to a solid phase.

2.9 Mineralogical Characterization

[0152] The Panalytical Empyrean x-ray diffractometer (XRD) was used under the following conditions: slit fixed at 10 mm, Cu/Kα monochromatic radiation, 40 mA and 45 kV. Samples were run at a speed of 2°θ/min(5-70°) to analyse the precipitates formed. Interpretation of data was done by the Highscore program. Samples were milled previously to a particle size less than 10 micron. Due to the small quantity of sample, a zero-background wafer sample holder was used.

[0153] The scanning electron microscope Jeol GSM 6610 equipped with energy dispersive system (SEM-EDS) was used for the analysis, along with Astimex 53 Minerals Mount MINM25-53 standards. The accelerating voltage of the beam during analysis was 20.0 kV with a spot size of 50 and working distance of 10 mm. Sample preparation for this method involves a strip of double-sided carbon tape attached to a glass section. The samples were coated with a thin layer of carbon (±15-100 nm) to prevent charging.

2.10 Results

[0154] FIG. 5 indicates that the data of the three columns are similar; the pH increased from 7.5 to 9.8 and the rest of the parameters (TDS, Sal, Cond, Eh, sulfates and iron) decreased from the first sampling performed after 24 hours. Fe.sup.2+ and Fe.sub.Total concentrations were always below detection limit (BDL). The sulfate concentration decreased from 1400 ring/L to BDL after one week. However Sal, TDS and Cond reached the lowest values after four months (from 980 mg/L, 1000 mg/L and 2090 μS/cm to 283 mg/L, 209 mg/L and 576 μS/cm, respectively). The Eh decreased from 296 to 150 mV within the first 24 hours and continued to decrease for two month to −21 mV, thereafter stabilized at about −35±15 mV.

[0155] The ICP analysis (Table 2) shows that the concentration of the Ba in the water increased in the first sampling, this is probably because the BaCO.sub.3 powder that is not attached to the wood shavings is released into the water, however the Ba concentration decreased and stabilized around 0.7 mg/L thereafter. Most of the compounds started to decrease within 24 hours such as Ca, Mg, CI, NO.sub.3, SO.sub.4 and Zn; from 262.4, 132.6, 9.2, 3032, 1253, 0.007 mg/L to 36.8, 97.8, 4.1, 1766, 147 and 0.003 ring/L, respectively (calculated as the average of the three columns). The rest of the compounds, such as Na, K, Al, Fe and Mn clearly started to be removed from the second sampling (5.sup.th week) from 4.9, 5.9, 0.044, 0.057 and 0.03 mg/L to 3.1, 5.3, 0.03, 0.008 and 0.002 mg/L, respectively (calculated as the average of the three columns). At the end of the experiment all the compounds were within the limits allowable for drinking water according to SANS 241 (South African National Standard 2006; 2011), except for the Mg that exceed the limit by 15 mg/L. The final removal of each compound is shown in Table 3. The similar evolution of the three columns, allowed for the spatial evolution analysis to be performed in column B and the geochemical characterization of the precipitates in column C.

TABLE-US-00002 TABLE 2 ICP water analysis of the main chemical compounds at the inlet and outlet of the columns A, B and C (As, Cu, Cd, Ni, Pb and Cr were always below detection limit (BDL)); NO.sub.3 Ca SO.sub.4 Mn Na Fe Al Zn Mg K Cl Ba INLET 3032 262.4 1253 0.03 4.9 0.057 0.044 0.007 132.6 5.9 9.2 0.11 Day 1 Sept A 1784 45.6 233.3 0.065 6.9 0.071 0.044 0.004 89.7 23.6 4.3 92.79 Sept B 1843 27.1 133.5 0.034 5.2 0.009 0.027 0.003 112.9 32.7 4.1 76.37 Sept C 1670 37.8 74.4 0.042 7.2 0.027 0.041 0.003 90.9 12.8 4 77.07 Month 2 Oct A 1685 11.2 240 0.002 2.9 0.008 0.029 0.006 132.3 4.6 5.9 0.95 Oct B 1701 9.7 218.2 0.002 3.2 0.011 0.029 0.004 137.8 5.2 5.6 1.36 Oct C 1715 10.6 253.1 0.003 3.1 0.004 0.031 0.004 133.2 6.1 5.4 1.45 Month 4 Dec A 2.8 5.3 116 0.002 1.4 0.024 0.006 0.002 118.2 6.7 9.6 0.67 Dec B 1 5.3 71.9 0.001 1.8 BDL 0.002 0.004 141.8 3.7 6.3 0.80 Dec C 1.1 4.8 68.6 0.001 1.2 BDL 0.005 0.003 118 4 6.4 0.73 Month 6 Feb A 0.3 4.7 78.8 0.002 BDL 0.006 0.008 0.002 111.1 4.8 6.6 0.75 Feb B 0.3 5 88.6 0.001 1.2 0.006 0.007 0.003 120 4.9 7.5 0.73 Feb C 0.3 5 96.6 0.003 BDL 0.055 0.009 0.003 113.2 11.8 14.8 0.74

TABLE-US-00003 TABLE 3 percentage removal of the main compounds in the three columns at the end of the experiment; Remov- al % NO.sub.3 Ca SO.sub.4 Mn Na Fe Al Zn Mg Column 99.99 98.2 93.7 94.2 100.0 89.3 82.0 73.1 16.2 A Column 99.99 98.1 92.9 95.6 74.3 89.4 84.6 65.9 9.5 B Column 99.99 98.1 92.3 89.0 100.0 92.2 78.8 65.8 14.6 C AVER- 99.99 98.1 93.0 92.9 91.4 90.3 81.8 68.3 13.4 AGE

2.10.1 Spatial Evolution

[0156] Two sets of samples from the four sampling ports, named from top to bottom: one, two, three and four, were collected from column B and analysed. The residence time of the MD in the column was 24 h, therefore the contact time with the reactive material of the samples from each port was approximately 6, 12, 18 and 24 h, respectively. The results show that the removal of every compound analysed occurred mainly at port one. The composition of the water at port two, three and four had no significant differences. Therefore, the fast dissolution of the BaCO.sub.3 in contact with the MD is displayed. This is also demonstrated by the analysis of precipitates.

2.10.2 Geochemical Modelling

[0157] The simulation was based on the physicochemical characteristics of the MD used as the solution in this experiment. Witherite was assumed as equilibrium phase 0, since total dissolution was expected. The predicted precipitates were barite (BaSO.sub.4; SI-3.49), calcite (CaCO.sub.3; SI-1.25), dolomite (CaMg(CO.sub.3).sub.2; SI˜2.50), Fe(OH).sub.3(a) (SI˜1.75), hausmannite (Mn.sub.3O.sub.4: SI˜15.34) and pyrolusite (MnO.sub.2; SI˜9.85). However, due the low concentration of Fe and Mn, those precipitates could be masked in the XRD analysis.

2.10.3 Mineralogical Characterization

[0158] According to the SEM-EDS analysis of the bottom sample, the composition of most of the crystals were mainly Ba (79-95%), O (4-21%) and some of them also had trace amounts of sulfur (0.5-4%) in the form of clear needles smaller than 5 μm. The XRD analyses determined that those crystals were 95.1% witherite (BaCO.sub.3) and 4.9% barite (BaSO.sub.4) (red diffractogram in FIG. 7). In the middle sample the concentration of witherite was lower (24.4%); the precipitation of barite increased (56.4%); the precipitation of Ca detected in SEM analysis was confirmed by the XRD analysis where calcite and aragonite where detected (6.4 and 12.8% respectively). However, the precipitation of Mg and K detected by SEM were masked in the XRD analysis, mainly due to the high concentration of barite (green diffractogram in FIG. 7). In the top sample the concentration of witherite and barite were slightly lower (23.4 and 46.7%, respectively), but calcite and aragonite concentrations were higher (7.6 and 22.3%, respectively). In this section of the column precipitates with Al, Fe, Mg, Na, Si, Cl and K were also found.

[0159] According to the analysis, the MD dissolves the BaCO.sub.3 in the top of the column and releases Ba.sup.2+ and HCO.sub.3.sup.−, both precipitate mainly as BaSO.sub.4 and CaCO.sub.3. Thereafter, the MD was already remediated and did not continue to react with the BaCO.sub.3 in the column. This is confirmed by the neoformed minerals found at the top and the middle, whereas the bottom sample still had reactive BaCO.sub.3 and no neoformed mineral phase was found. Furthermore, in the picture of FIG. 7, it can be observed that the bottom of the column is still white due the BaCO.sub.3 that remained on it.

2.11 Conclusion

[0160] 764 L of alkaline coal mine drainage from the site was treated by the B-DAS (Barium carbonate-dispersed alkaline substrate) system in lab scale bioreactors. The aim to remove the high cations and anions concentration as well as the Sal and TDS from this drainage was achieved. According to the water analysis and the mineralogical characterization, the B-DAS system has demonstrated the capacity to remove 93% of sulfates through the precipitation of barite (BaSO.sub.4); 98% of Ca by precipitation of calcite and aragonite (CaCO.sub.3); remove Mn, Na, Fe, Al, Zn, Mg (93, 91, 90, 82, 68 and 13%, respectively). K and Si were also found in the neoformed precipitates. NO.sub.3 was also removed (99.9%) from the MD, but the absence of N in the precipitates and the extremely reductive condition in the bioreactor (Eh about 35 mV) could have promoted the denitrification process. The EC, Sal and TDS decreased about 50-70%.

[0161] According to the XRD analysis, after 6 months, column C had about 22% of the BaCO.sub.3 at the top and 95% at the bottom of the column. Therefore, the reactive capacity of the BaCO.sub.3 could be extended. Neoformed crystals were found in the top and middle samples, but not in the bottom sample, indicating that the dissolution of the BaCO.sub.3 and the consequent precipitations took place in less than six hours (estimated residence time of the water in the top section of the column), demonstrating the effective treatment and the capacity of this system.

Nitrate Bioreactor

2.12 Sampling

[0162] Samples were taken at the K1 Return Dam (Mine), Kroondal on the 21 Nov. 2014 to perform various tests and experiments on. This mine is a significant primary producer of the platinum group metals (PGMs), which comprise platinum (Pt), palladium (Pd), rhodium (Rh), osmium (Os), ruthenium (Ru) and iridium (Ir). The used mine water is pumped into a return dam and it is reported that sewerage leaches into this dam.

2.13 Water Quality

[0163] The water samples were sent to the Institute for Ground Water Studies (IGS) at the University of the Free State to analyse the chemical parameters of the on-site water. The IGS chemical data is presented in Table 4. The chemical compounds indicated in red are over the allowable SANS level for class 1 drinking water.

TABLE-US-00004 TABLE 4 IGS water quality results from K1 Return Dam at Kroondal; Determinant K1 Dam SANS* pH 7.16 >5.5 ORP 69 — Electrical Conductivity (mS/m) 244 170 Salinity (mg/l) 129 — Total Dissolved Solids (mg/l) 1730 1200 Fluoride as F (mg/l) BDL* 2 Bromide as Br (mg/l) BDL* — Chloride as Cl (mg/l) 184 300 Nitrate as N (mg/l) 123 11 Nitrite as N0.sub.2 (mg/l) 15 11 Total Ammonia as N (mg/l) 26 1.5 Phosphate as PO.sub.4 (mg/l) 1 15 Sulfate (mg/l) 305 500 *SANS—South African National Standards 241: 2006&2011 for drinking water class 1 *BDL—Below Detection Limit

2.14 Organic Pollutant Load (On-Site Water)

[0164] A Biological Oxygen Demand (BOD) test determines the amount of Dissolved Oxygen (DO) which indigenous microorganisms take up to break down organic material present in water to grow over a period of 5 days. The BOD from K1 Return Dam (Aquarius SA) was as follows:

[0165] Before treatment: BOD mg/L=33 mg/L (DOC equivalent of 31 mg/L)

[0166] A relatively high 33 mg/L BOD indicates two aspects: [0167] There is a high level of microbial activity and thus counts in the water, [0168] There is an organic donor source, usually considered as contaminants. This confirms the reported sewerage leaching into the return dam.

[0169] E. coli plate counts as well as Total Coliforms counts were performed on the on-site water sample. The Coliforms counted 1986 cfu/100 mL, where the maximum allowed by the SANS class 1 for drinking water is 10 cfu/100 mL. The E. coli tested at 921 cfu/100 mL, where SANS allows less than 1 cfu/100 mL for class 1 for drinking water. The Coliforms thus exceeds the allowed limits and this is also the case for E. coli. This bacterial load can however be used to bio-remediate the nitrate in the water if they possess the Nitrate Reductase Genes. The organic pollutant load (sewerage) can serve as a carbon source for denitrifying bacteria in the water to bio-remediate the nitrate pollution.

2.15 Biodiversity of the Indigenous Bacteria

[0170] A microbial community can be monitored over a period of time to determine the organisms present for nitrate reduction by using cell counts, denaturing gradient gel electrophoresis (DGGE) and sequence analysis. The electron acceptor, as well as the products formed, is monitored to determine if there is a correlation between the microbial community and nitrate reduction. The biodiversity analyses can further be extended from the regular 16S rDNA genes to functional gene markers for denitrifying bacteria using Nitrate Reductase Genes (nirK and nirS).sup.8.

[0171] The on-site water DGGE and sequencing results were inconclusive and could not identify most of the bacteria, probably due to high genetic contamination in the water. However, the sequencing did identify two organisms, namely Pseudomonas stutzeri and Flavobacterium sp. Pseudomonas stutzeri is a Gram-negative, rod-shaped, motile, single polar-flagellated bacterium found in soil. P. stutzeri has the ability to denitrify polluted water.sup.9 as literature indicates that it has Nitrate reductase and Denitrification Regulatory Protein nirQ.sup.13. Flavobacterium is a genus of Gram-negative, non-motile and motile, rod-shaped bacteria that consists of 130 recognized species. These bacteria are commonly found in soil and fresh water and cause disease in trout.sup.10. The DGGE and sequencing is being optimized and repeated. Based on the DGGE results the indigenous bacteria, especially Pseudomonas stutzeri, can be used in denitrifying conditions to remediate the nitrate levels.

2.16 Columns

[0172] Two denitrification experiments were conducted using K1 Return Dam water as influent. The experiment had two columns in series. The first column was constructed of 0.5 m PVC pipe with a 110 mm inner diameter with threaded end caps and with taps and appropriate rubber and silicone tubing. Influent water was delivered to the base of reactor using a Watson Marlow peristaltic pump. The reactor was packed with a dolerite matrix with 54.7% porosity. The working volume of the reactor is 1860 ml and was operated with a 1.3 ml/min flowrate to obtain a 24 Hydraulic Retention Time (HRT). BaCO.sub.3 was added as a pre-treatment to stabilize the redox chemistry by creating the correct ORP level for nitrate reduction. An up-flow column were constructed from PVC pipe (10 cm inner diameter, height of 50 cm) and equipped with two additional lateral samplings ports. Each port had a small perforated pipe extending into the column matrix to promote homogeneous samplings and allow for homogeneous flow within the columns, by increasing the area of sampling while, avoiding, as far as possible, preferential flow. Each column contained a layer of quartz gravel (particle size about 5-8 mm) at the bottom (2.5 cm). This layer was covered with a 40 cm reactive material layer, which consisted out of a BaCO.sub.3 and wood shaving mixture. The ratio of wood:BaCO.sub.3 (w/w) were 1:3. If the oxygen levels are low enough for long enough, the aerobic bacteria will die and will be out competed by the denitrification bacteria.

2.17 Carbon Source

[0173] The glycerol was selected as carbon source since is cheaper and effective to promote the growth bacterial. The volume added to the reactor was 0.69 mg/L/min.

2.18 Results

[0174] FIG. 5 illustrate the barium carbonate results. The denitrifying bacteria accordingly reduced the nitrate gradually to 95% on day 14 (FIGS. 8a and 8b). The nitrate is reduced to 5.5 mg/L on day 14 and sulfate reduction by 90% on day 14. These values are within the SANS limits for drinking water. The spike in dissolved oxygen on day 9 (FIGS. 8a and b) had no effect on the nitrate reduction, which indicates that the dominant bacterial specie is therefore denitrifying bacteria.

Cynaide Bioreactor

2.19 Starting Materials

[0175] Water samples (about 25 L) were each collected into polyethylene carboys on 2 Dec. 2013 from two sites metallurgy plant: (1) from T #1 Process Tank which was aerated and lime was added to maintain an alkaline pH, and (2) from L dam whose hydrogeochemical characteristics were typical of AMD water. The samples were to be used as feedstock in preliminary lab-scale remediation experiments. The carboys were stored at 4° C. until use.

[0176] Samples (50-500 ml) were also taken from both sites as well as the as the clarifier tanks and filtered through a 0.45 μm Teflon filter within 24 h of collection for chemical analysis.

2.20 Bioreactor Columns

[0177] Two DAS bioreactors connected in series were used to treatment of MDs. Due to the low pH, L dam water was treated with twin DAS bioreactor. While, a DAS bioreactor was bio-augmented with an inoculum of sulfate reducing bacteria, followed for an activated charcoal cartridge and DAS bioreactor were used for the T water sample. Samples of effluents from each reactor were collected after 24, 72 and 120 hours for analysis.

[0178] The DAS bioreactor used in this study were similar to the bioreactors already descripted above. The bioreactor that was bio-augmented had similar appearance and configuration as the DAS bioreactor with 4 (27.5 cm distance) and 7 (13 cm distance) lateral portals on either side of the column. The main differences were that this column had 1 meter height and contained a mixture of 60% wood chips 20% BaCO.sub.3, and 20% bacterial inoculum. The space occupied by these materials in the column was 80 cm and it is used in down-flow bioreactors. The porosity of the systems was volumetrically calculated. The column was filled with sample water to cover the reactive material and then the volume of water that was used was measured. The initial flow rate for either column was 1.09 ml/min (1.4 L/day) with a residence time of 24 hours for the DAS bioreactor and 24 hours for bioreactor bio-augmented. However, the flow was increased to 5 L/day throughout the experiment.

2.21 Results

[0179] 2.21.1 Characterization of Water Samples from #1 T Process Plant and L Dam

[0180] The initial parameters of #1 T process tank and the L dam, determined using hand-held probes as well as laboratory assays are displayed in table 5. As expected, the T water had an alkaline pH due to the addition of lime. The water sample also contained 436 μg/L CN.sub.WAD and approximately 1800 mg/L SO.sub.4.sup.2−. On the other hand, the L AMD water was acidic, contained approximately 164 μg/L CN.sub.WAD, 8100 mg/L SO.sub.4.sup.2− and 504 mg/L Fe.sub.Total.

TABLE-US-00005 TABLE 5 Parameters of water samples from #1 T process tank and L dam; Parameter #1 T process tank L Dam pH 8.5 2.35 Conductivity (mS/cm) 3.57 8.40 Salinity (mg/L) 1800 3770 TDS (mg/L) 2470 6000 SO.sub.4.sup.2− (mg/L) 1800 8100 Cyanide.sub.WAD (μg/L) 436 164 Fe.sub.Total (mg/L) 2 504 Fe.sup.2+ (mg/L) 2 12

2.21.2 #1 T Process Tank (DAS Bioreactor Bio-Augmented (SRB-DAS) and DAS Bioreactor (BDAS))

[0181] The process flow for treatment of the T water sample was as follows: Sulfate reducing bacteria (SRB) DAS.fwdarw.activated charcoal cartridge.fwdarw.DAS bioreactor (BDAS).fwdarw.final effluent. This flow direction was employed because the pH of the water (8.5) was within the limits tolerated by the SRB (pH 5-9). Moreover, the pH is within the bioreactor is stable since the precipitation of metal sulfides generate acid or release hydrogen ions which decrease the pH minimally to around pH 6-7, enough to generate optimal growth conditions for SRB. The activated charcoal cartridge was included to investigate its effect on the removal of the odour associated with H.sub.2S production as well as the removal of contaminants.

[0182] The efficiency of the SRB-DAS was investigated for SO.sub.4 removal from the T water feed. After 24 hours of treatment, SO.sub.4 was not detected in the samples collected, indicating a 100% removal. A minimal SO.sub.4 concentration of 6 mg/L was measured after 120 h; however, this residual SO.sub.4 was mopped up after passing through the DAS column. No sulfate was detected in the final sample collected from the DAS column at the end of the experiment (Table 6).

[0183] Approximately 30% of the cyanide concentration in the sample was degraded with the first 24 hours of treatment with the SRB-DAS and, a further decrease of up to 42% was observed after flowing through the cartridge and BDAS. This significant decrease suggested that the microbial consortium in the bioreactors were capable not only of sulfate removal but also cyanide degradation. The observation that cyanide degradation proceeded faster in the SRB-DAS compared to the DAS bioreactor was a possible indication that the microorganisms were also involved in cyanide degradation improving the prospect of investigating larger volumes and higher water flow rates. After 120 hours, the cyanide concentration in the final effluent had decreased from 436 μg/L to 41 μg/L; below SANS recommended levels (FIG. 9). These values obtained corresponded to a 90.6% removal of cyanide from the T water sample.

[0184] The ferrous and total iron concentrations decreased up to 0.12 mg/L, which were below the SANS recommended standards for drinking water. Moreover, there was a decrease of approximately 50% in the measurements of electrical conductivity, salinity, and TDS. Although the results were slightly higher than the recommended levels, the trend observed after 120 hours showed a steady decrease in the measured values which might indicate a possibility of the parameters reaching acceptable levels. However, we were unable to obtain more data points as the experiments had to be halted due to the limited water sample available.

TABLE-US-00006 TABLE 6 Parameters measured after 120 hours remediation of water from #1 T process tank, compared with SANS standards. The values highlighted in green were within the limits of SANS recommended levels; #1 T SRB- SANS recommended process DAS + levels for drinking Parameter tank BDAS water pH 8.5 6.7  5.5-9.7 Conductivity (mS/cm) 3.57 1.87 ≦1.70 Salinity (mg/L) 1800 927    — TDS (mg/L) 2470 1304     ≦1200 SO.sub.4.sup.2− (mg/L) 1800 0*   ≦500 Cyanide.sub.WAD (μg/L) 436 41    ≦70 Fe.sub.Total (mg/L) 2 0.12 ≦2 Fe.sup.2+ (mg/L) 2 0.12 — *SO.sub.4.sup.2− had been completely removed from the effluent after 24 hours.

2.21.4 L Dam (Twin BDAS)

[0185] The evolution of the L dam water sample after 120 h is shown in Table 3. The results obtained showed that the twin BDAS were very effective in removing SO.sub.4.sup.2−, Fe.sup.2+ and Fe.sub.(Total), as well as neutralizing the pH and adjusting the conductivity, TDS and salinity closer to SANS acceptable standards. Complete removal of SO.sub.4.sup.2− and 99% removal of Fe.sub.(Total) were achieved within 24 h of water flow through the first BDAS column. In addition, there was at least a 50% decrease in conductivity and TDS after the first bioreactor and 65% after the bioreactor column. These parameters were less than twice the acceptable limit compared to the original sample which was 5 times the acceptable limit (Table 7). The cyanide concentration decreased by 69.4% from 164 μg/L to 50 μg/L, below recommended levels.

[0186] Although the bulk of contaminant removal took place in the first bioreactor, the attachment of a second bioreactor allows for an increase in water flow rate meaning more volumes of water can be treated while the longevity of both bioreactors is extended. Based on results from previous experiments where AMD with similar characteristics to the L water were evaluated for 5 months, our treatment is stable and will yield similar results in pilot scale treatment processes.

TABLE-US-00007 TABLE 7 Evolution of L Dam AMD sample after 120 hours remediation with BaCO.sub.3-DAS columns whereby the values highlighted were within the limits of SANS recommended levels; BDAS BDAS SANS levels Bioreactor Bioreactor for drinking Parameter L Dam #1 #2 water pH 2.35 6.83 7.09 5.5-9.7 Conductivity 8.40 3.58 2.94 ≦1.70 (mS/cm) Salinity (mg/L) 3770 1750     1440     — TDS (mg/L) 6000 2470     2050     ≦1200 SO.sub.4.sup.2− (mg/L) 8100 0*   0*   ≦500 Cyanide.sub.WAD (μg/L) 164 59    50    ≦70 Fe.sub.Total (mg/L) 504 0.51 0.84 ≦2 Fe.sup.2+ (mg/L) 12 0.45 0.3  — *SO.sub.4.sup.2− had been completely removed from the effluent after 24 hours.

Phase 3: Pilot Scale of BDAS System (Alkaline MD Bioremediation Construction and Running)

3.1 Installation and Construction

[0187] The environmental impact of mining was evaluated for 5 days. In this visit, a scientific and engineering team firstly located the main sources of pollution and conducted a mapping of the area in order to know the characteristics of the terrain. This helped us to plan the distribution of the treatment system in the selected location.

[0188] The starting alkaline materials, such as BaCO.sub.3, were purchased from Protea Chemicals Company. BaCO.sub.3 have a purity of 88.6% according to Protea Chemicals. Moreover, these materials contain major metals such as Fe and SO.sub.4.sup.2− but in negligible concentrations.

[0189] The environmental media contaminated with a source of MD crosses two decanters in a residence time of 12 hours and finally is evacuated in a unique flow rate of 4.8 L/min towards the aeration system (cascade). The decanters play a vital role in the precipitation of salts, TDS and solids suspended.

[0190] FIG. 10 shows the cascade system that the reaction occurring in the reactor consumes oxygen. Therefore, an aeration system is needed to recover the lost oxygen and to stabilize the pH. This system consists of an open channel cascade that allows the fall of water. Along this cascade, bubbles are formed allowing environmental media contaminated with a source of MD to interact with oxygen and CO.sub.2 to acidify the environmental media contaminated with a source of MD and stabilize its pH. In addition, a second site visit was used to install PLC systems to measure the pH and Conductivity of the system at various stages. FIG. 11 shows the integration of the PLC system with the whole pilot plant.

3.2 Results

[0191] 2 592 000 L of alkaline coal mine drainage from the site was treated by the B-DAS (Barium carbonate-dispersed alkaline substrate) system in pilot scale. Bioreactors with a resident time of 24 h were working from July (tank 3) and August (tank 1, 2 and 4) until April. Four tanks were installed at the mine facility for the treatment of MD stored in a dam. The main difference between the tanks was the proportion used in the mixture of reactive material (BaCO.sub.3) and wood chips and the flow of each tank. The ratio was 1:1 for tank 3 and 1:3 for the other tanks (1, 2 and 4). Tank 1 was selected to represent tank 2 and 4 for the chemical analysis. Tank 1 had a flow rate about 1.2 L/min. While, the flow in tank 2, 3 and 4 was increased up to 2.5 L/min after 3 months, to promote the fast saturation of the system. The system ran properly during the 7 months of testing. The system seemed to be saturated by March as the system was losing reactivity at that stage.

[0192] The main aim of the aforementioned system was achieved; being the removal of high cations and anions concentration, as well as the Hardness, Salinity (Na.sup.+, Cl.sup.−, SO.sub.4.sup.−2 and Ca.sup.2+) and TDS from the environmental media contaminated with a source of mine drainage (MD). The chemical processes involved in the anions (SO.sub.4.sup.−2) and cations (Mn.sup.2+, Fe.sup.2+, Al.sup.3+, Zn.sup.2+, Ca.sup.2+, Mg.sup.2+ and Na.sup.+) removal are described in the equations 15 to 21.

[0193] According to the water analysis, the B-DAS system has demonstrated the capacity to remove between 76% (tank 1) and 53% (tank 3) of sulfates (during the first 7 months) through the precipitation of barite (BaSO.sub.4). In tank 1, the sulfate values (341.51 mg/L on average) were always inside the limits of SANS drinking water standard for class 1. While in tank 3 and cascade the values were over the limits of SANS drinking water standard for class 1, when the flow was increased. It is noteworthy that the values in the cascade were increased due to improper functioning of tank 3 and the rapid saturation of tanks 2, 3 and 4. To prevent it, the flow of tanks that were refilled was not increased.

[0194] The removal of Ca was among 95 and 88% for tanks 1 and 3, respectively; in the cascade was about 91%. Also, high percentages of removal for CI and Na can be seen. In tank 1 and 3, the CI and Na % removal was about 59 and 26%, respectively. While, in the cascade the removal decreased to 23 and 16%, respectively.

[0195] The metals contained in this drainage were completely removed. The Ba values (0.1 mg/L) did not exceed the limits of SANS drinking water standard for class 1 (0.7 mg/L). The Mg concentration was slightly removed in tank 1, but not removal was observed in the tank 3 or cascade.

[0196] However, the mineralogy characterization by SEM-EDS showed the presence of crystalline neoformed phase minerals of Mg. As well as, probably, crystalline neoformed phase minerals of calcite and aragonite (CaCO.sub.3) and Na/CI complex.

[0197] The EC and TDS decreased 52%. Specifically, the TDS was removed due to the precipitation of SO.sub.4.sup.2− as barite (BaSO.sub.4) and cations such as Ca, Na and CI, mainly. The precipitation of these neoformed mineral and metallic phases allowed also the decrease of EC and TDS.

[0198] According to the thermodynamic modelling; after 6 months, the tank should be saturated and has low reactivity. The results from the chemical data demonstrates that the treated drainage is supersaturated with respect to barite (BaSO.sub.4; IS˜3.49), Calcite (CaCO.sub.3; IS˜1.25), dolomite (CaMg(CO.sub.3).sub.2; IS ˜2.50), Fe(OH).sub.3(a) (IS˜1.75), Hausmannite (Mn.sub.3O.sub.4: IS˜15.34), Pyrolusite (MnO.sub.2; IS˜9.85). According to the thermodynamic model when the water comes into contact with a solid material (e.g. some mineral or rock) and exceed equilibrium with that material these minerals phases are generated, which would be the most thermodynamically stable phase in our system.

[0199] Tank 1 was the most effective. The improvement of the water characteristics reached values within the limits of SANS drinking water standard for class 1, except for Mg (FIGS. 30 and 31). These results demonstrate that a semi-passive treatment based on BaCO.sub.3 (BDAS) could be a viable solution in comparison with other systems whose costs are excessive especially for the treatment of closed mines. After 10 months the reactive material was removed from the tank and properly stored. Then, 2 reactive tanks were filled with new alkaline material following the methodology carried out in the phase 2.

3.3 Sampling and Analysis of Reactive Material

[0200] After 10 months running, the reactive tanks were stopped and drained. Then, the reactive material was excavated and samples were extracted at different depths. The samples were collected in plastic bags and labelled properly. Finally, the samples were stored and transported to the lab where they were dried and analysed by SEM-EDS (FIG. 13) (XRD in progress). The analysis by SEM-EDS showed the presence of crystalline neoformed phase minerals that contains mainly Ca, Ba, SO.sub.4.sup.2− and Mg. As well as Na/CI complex poorly crystallized and distributed over the calcite or aragonite. The main crystalline neoformed phase minerals obtained were barium sulfate (barite) and calcium carbonate (aragonite and calcite).

3.4 Toxicity Analysis

[0201]

TABLE-US-00008 TABLE 8 Test results and risk classification of B-DAS system; Results Wastes Exxaro Exxaro Leachate inlet outlet pH 7.3 6.7 8.2 EC(Electrical conductivity) 9.1 204.9 115.9 (mS/m) Disosolved oxygen (mg/l) 9.3 9.1 5.1 V. fischeri Test started on 15-07-13 15-07-13 15-07-13 (bacteria) yy/mm/dd %30 min inhibition 7 15 −39 (−)/stimulation (+) (%) EC/LC20(30 mins) n.r. n.r. 41 EC/LC50(30 mins) n.r. n.r. n.r. Toxicity unit (TU)/ <1 <1 <1 Description P. reticulata Test started on 15-07-16 15-07-16 15-07-16 (guppy) yy/mm/dd %96 hour mortality 0 0 0 rate (−%) EC/LC10(96 hours) n.r. n.r. n.r. EC/LC50(96 hours) n.r. n.r. n.r. Toxicity unit (TU)/ <1 <1 <1 Description Estimated safe dilution 100 100 41 factor (%) [for definitive testing only] Overall classification - Hazard class*** Weight (%) Class II - Class I - No Class I - No Slight acute/chronic acute/chronic acute/chronic hazard hazard hazard 0 0 50

[0202] From Table 8 it is clear that samples EWL, EXI showed “no acute/chronic toxicity hazard” (Class I). However, samples EXO showed “slight acute/chronic toxicity hazard” with safe dilution factors calculated of 41 (41 parts source water with 59 part unpolluted water). Therefore the leachates that could be produced during collection and storage of the waste produced by the water treatment are no toxic and no special storage would be required. The drainage collected from the dam was no toxic for the bacteria nor for the fishes. After the treatment, the drainage showed no toxicity for the fishes, but slight toxicity for the bacteria (probably due to the sensibility of these specific bacteria to the wood resins).

3.4 Emptied and Filled

[0203] The starting materials were placed next the system to facilities its handling during the filling of the tanks. Due to lack of TLB or crane, none of the containers were replaced and only 2 tanks were manually refilled (tank 1 and 4).

[0204] In tank 1; the old material was removed and replaced with the new mixture. While in tank 4 only part of the old material was removed and the new mixture was placed on top. The proportion used in this mix was 150 kg of BaCO.sub.3 and 120 kg of wet wood chips. The mixing was made manually. The tanks were emptied and filled manually, as well.

[0205] Finally, it was made maintenance of the cascade, change taps and the flow rate was regulated to 1.5 L/min for each tank. The sampling was carried out on next day.

[0206] After 5 months of replacement, the results are encouraging both in the tanks as in the cascade, with sulfate removal rate 78% (from 1373 to 296 mg/L) on average and about 64% removal of TDS (from 1050 to 380.9 mg/L) and thus, removal of hardness from waters. Also, levels of pH values about 9.13 on average and conductivity values averaged 0.65 μS/m. All of them within the limits for drinking water (FIG. 14a, 14b and FIGS. 15a to 15e). There were no significant differences between the tank totally refilled and the tank where BaCO.sub.3 was stacked. However, it is recommended to empty and refill the tanks to prevent an increase of anaerobic zone. After 15 months in total, we can say that this is the right system for the treatment of water with high concentration of sulfates, TDS and EC, so far. Also, no system has come so far, generating a by-product that can be recycled as BaSO.sub.4 and CaCO.sub.3 (no toxic by-product) and fully decontaminated water.

3.5 Lab and Pilot Scale of BDAS System for Treatment of Extremely Acidic Drainage

[0207] The drainage treated in this case is leaching from phosphogypsum stacks. Phosphogypsum refers to the gypsum formed as a waste/byproduct of processing phosphate rock into fertilizer through a wet chemical process (equation 33).

Ca.sub.3(PO.sub.4).sub.3F+5H.sub.2SO.sub.4+10H.sub.2O.fwdarw.3H.sub.3PO.sub.4+5CaSO.sub.4.2H.sub.2O+HF  (33)

[0208] The action of sulphuric acid (H.sub.2SO.sub.4) on phosphate rock, mainly fluorapatite (Ca.sub.5(PO.sub.4).sub.3F), yields phosphoric acid (H.sub.3PO.sub.4), hydrogen fluoride (HF) and gypsum (CaSO.sub.4*2H.sub.2O). The wet method generates about 5 ton of waste, commonly named phosphogypsum, per ton of phosphoric acid manufactured. These wastes are highly enriched in metal impurities, radionuclides from U-decay series and rare earth (5% de lanthanides in the case under study). Phosphogypsum wastes are often disposed in large stockpiles exposed to weathering processes, where they may cause serious environmental damage.

[0209] The stack under study contains about 38 MTones of phosphogypsum, its mineralogical composition is: brushite (>50%), gypsum (20-10%), apatite (<2%) and clay (<2%); chemically, it is worth noting the following trace elements (sort from largest to smallest): Sr, Th, Ba, Mn, Y, As, Cu, Sc, Se, U, Ag, Mo, Zr, Cd, Cr, Pb, Tl, V, Ni, Zn, Sb, Co and Br. Most of these metals are toxic or radioactive and they are leachates to the environment.

[0210] The infiltration of the leachates from the gypsum stacks during the last 50 years has been accumulated in the shallow weathered zone aquifer, creating a contamination plume around the gypsum stack and the impoundment dams. The groundwater has been impacted by acid containing materials manifest as a low pH and high Total Dissolved Solids (TDS). Often the TDS is predominantly associated with sulfate salts. There are also some instances where the metal concentrations in the groundwater (aluminium, iron, manganese, arsenic and copper) are out of compliance against the SANS 241:2006 Drinking water standard. Presumably, below the gypsum dams the SO4 concentration in the aquifer is above 5000 mg/l. The gypsum leachates are partially collected in impoundment dams and pumped back to the top of the stacks. The natural and forced evaporation processes that take place in the dams over the past 50 years have contributed to increase the concentration of pollutant over time.

[0211] For the remediation of the drainage, water samples from John's Dam (JD) was collected and treated with BDAS system at lab scale bioreactor. Thereafter, a bioreactor at pilot scale was designed and installed on site to treat the drainages collected at JD. In this particular case, the drainage was treated with a combination of CaCO.sub.3-DAS and B-DAS, with the aim of increasing the pH and therefore decreasing the consumption of the BaCO.sub.3.

3.6 Column Experiment

[0212] Two down-flow columns connected were constructed from PVC pipes (10 cm inner diameter, height 50 cm) and equipped with four additional lateral sampling ports.

[0213] Each column contained a layer of quartz gravel (particle size about 5-8 mm) at the bottom (2.5 cm). This layer was covered with a 40 cm reactive material layer, which consisted of CaCO.sub.3 and wood shaving mixture in the first column (column A) and BaCO.sub.3 and wood shaving in the second column (column B). The ratio wood: reactive material was 1:3 (w:w) in both columns 600 g of reactive material and 200 g of wood shaving.

[0214] The contaminated water is pumped to column A, thereafter it flows gravitationally through the column and it is pumped to the top of column B. The outflow from column B was collected in a container that also functioned as an aeration and settling tank. The flow rate was 1 mL/min with a residence time of 24 hours for each B-DAS columns. The porosity of the systems was 70% (volumetrically calculated).

[0215] Water analyses performed as described previously in phase 1 and 2.

3.7 Results

[0216] The physicochemical parameters measured show a clear improvement in the characteristics of the drainage treated (it is worth to noting that the concentration of pollutants of this particular drainage is extremely high); the acidity is neutralized (pH increased from about 3 to about 7), conductivity, salinity and total dissolved solids decreased 39.1, 39.9 and 36%, respectively; The concentration of sulfate decreased 85.8%.

TABLE-US-00009 TABLE 9 Physics-Chemicals parameters Inlet and outlet; CaCO3 Output start 18 Feb column CaCO3 + BaCO3 parameter Units blank 24-Feb 24-Feb Temperature ° C. 23.7 23.1 22.7 Atmosf. Pres mm Hg 645.1 643.2 643.7 DO % 42.6 35.3 5.5 DO mg/L 2.93 2.45 2.91 SPC μS/cm 13792 13161 10058 Conductivity μS/cm 13439 12679 8410 Resistivity Ω cm 74.41 78.87 103.64 TDS mg/L 8963.5 8554 5740 Salinity g/L 7.97 7.58 4.87 pH 3.03 5.12 7.06 ORP mV 342 259.7 127 SO4 mg/L 7750 1100

3.8 Bioreactor at Pilot Scale

[0217] Due the poor quality of limestone (CaCO.sub.3) and its heterogeneous grain size, the amount of water that could be treated by the big limestone tank (BT) is less than the original design; therefore another reservoir tank (RT) was necessary to ensure the continuous flow to four barium reactors (BR) and the capacity of BT was not enough to sustain more than 4 BRs.

[0218] The water is pumped from JD to BT, which has 30 m3 of volume with 15% of porosity, therefore it has about 5000 L of capacity plus 2000 L of supernatant. The water is recirculating into BT until pH-meter get a value of 4.0. Then, the electronic valve releases the water to reservoir tank. Before that BT is completely empty, the pump installed in JD start to pump again into BT.

[0219] There are two reservoir tank (RT) of 5000 L onto of the 3 m stand, they have three objectives: [0220] To let the system run without any extra energy, the water flows by gravity from RT into barium carbonate reactors (BR); [0221] It also keeps the parameters stabilized, so we can control the quality of the water that is going into BRs; and [0222] Finally it ensures continuous flow into BR, therefore we avoid water stagnation and we can control the residence time into each BR.

[0223] There are 4 down flow bioreactors, 5000 L per tank, called BR1, BR2, BR3 and BR4. Each reactor have a manifold at the bottom to avoid preference flow within the tank. The bottom of each tank is filled with gravel up to 10 cm over the manifold. The gravel is used as a filter and also contributes to avoid preferential flow.

[0224] The rest of the tank is filled with a mixture of barium carbonate (BaCO.sub.3) and wet wood chips. We have been bench marking in lab scale, several ratios of Barium: wood chips and with these results we concluded that a 2:1 ratio was the best option. Therefore each tank is filled with about 625 Kg of BaCO3 and 312 Kg of wood chips. The wood chips are inert material used as a matrix to support the BaCO3 powder that is adhered to the wood chips and it also gives the adequate porosity to each reactor (about 50%). Each tank has a capacity of 2.5 m.sup.3, in order to get a residence time about 12 hours the flow rate was 180 L/h per tank.

[0225] The decanter was an 8 m.sup.3 square tank with a stilling wall disposed perpendicular to the flow direction, therefore there are two sections; the turbulence flow section where the inlet is creating turbulence and the laminar flow section where the characteristics of the wall avoid any turbulence and allow the water flow to be slowed down and have laminar flow.

[0226] The objective of this tank is decreases the total suspended solids (TSS), in addition to keep the parameters stabilized and to avoid that any barium carbonate could be released.

[0227] The reactions that are taking place in BRs are oxygen consuming; therefore an aeration system (AS) was needed to recover the lost oxygen.

3.9 Sampling

[0228] Samples were collected periodically from JD, BT, RT, BRs, DT and AS (outlets of the pilot plant) with the aim of follow the evolution of the drainage through the pilot plant. Physicochemical parameters were measured on site, as describe before, and cations and anions concentration were analyzed by ICP.

3.10 Results

[0229] pH was increased from 2.3 up to 4 in BT and then the water is stored in RT to ensure continuous flow at BR, where pH increases up to 7. The pH keep stabilized from there to the outlet (through DT and AS). EC decrease 64.15%. Ca is increasing at BT because of the release of Ca from the lime (CaCO3), but in BR decreases 90%. Mg decrease about from 700 mg/L to 420 mg/L (40%) at BR and up to 85 mg/L (88%) at As. Na decreases 25% at BR, the main reduction is due to the aeration at AS (46%). F was removed 100%. CI was removed 23%. PO.sub.4.sup.3− was completely removed (from about 8000 to 0 mg/L). SO.sub.4.sup.2− was removed 77.5%, (from an average of 4237.5 to 995 mg/L). Al was removed 99.99%. Fe was removed 99.93%. Mn was removed 99.89%. BT was a reused tank that was contaminated with son compounds, that is the reason for the arsenic to appear at the analyses done in BT and followings but it was 100% removed at the outlet. Ba at inlet was 0.027 mg/L, but it was increased at the barium carbonate reactor (BR) due the release of Ba. Therefore the concentration of Ba have been increased 11.5%, but it is still under the allowable limit for any use of water (drinking water (0.7 mg/L), irrigation livestock, etc) 0.032 mg/L. Co and Cr were 100% removed. Cu was 99.7% removed. Ni was removed 98.4%. V was removed 99.6%. Zn was remove 98.1%. NO3, No2, Br and Pb were under detection limit since the inlet (JD) throughout the whole system.

3.11 Conclusion

[0230] The main removal of most of the compounds took place into the BR. Some compounds are increased at BT and/or DT. Both tanks were reused, therefore and even after the cleaning labor, some contaminants may remain in the tanks and may be released during the experiment. BT removed most of the Cu, Co, Cr, Zn, V and F, while BRs removed most of the SO4, PO4, Ca, Mn, Ni and 40% of Mg. DT is settling the Ba by decreasing its concentration from 0.08 to 0.01 mg/L. AS helped to remove Mg and Na (50% and 38%, respectively, of the total removed throughout the system), as well as Mn, CI, SO4, PO4 and EC.

TABLE-US-00010 TABLE 10 Example of the parameters analyzed during a routine sampling performed in April 2014; BT BR1 BV BT RT BV BR1 JD BV BT 240414 BV RT Top BV.JD.090414 BV.JD.240414 100414Right Pump 100414 110414 pH 2.31 2.41 5.34 4.08 4.25 7.33 EC 967 1142 712 826 779 525 Ca 389 417 460 520 601 63 Mg 634 715 601 710 719 441 Na 737 802 536 742 724 582 K 103 104 40 69 65 92 F 528.2 411.7 −10.0 −10.0 246.4 −1.0 Cl 682.6 697.2 599.1 696.5 706.5 645.8 NO2(N) 0.0 0.0 0.0 0.0 0.0 0.0 Br 0.0 0.0 0.0 0.0 0.0 0.0 NO3(N) 0.0 0.0 0.0 0.0 0.0 0.0 PO4 8351.1 7383.7 4931.7 6553.3 7004.5 669.8 SO4 4029.7 4445.4 3395.4 4397.4 3639.0 1370.9 Al 109.171 117.416 0.400 23.269 8.734 −0.001 Fe 24.651 17.668 0.022 0.079 0.067 0.165 Mn 16.282 18.079 6.436 12.634 12.008 1.218 As <0.006 <0.006 0.101 0.063 0.100 <0.006 Ba 0.018 0.038 <0.001 0.019 0.004 0.044 Co 0.243 0.273 0.052 0.108 0.087 <0.002 Cr 0.129 0.099 <0.006 0.011 <0.006 <0.006 Cu 3.823 3.731 0.118 0.492 0.232 0.006 Ni 0.851 0.654 0.325 0.446 0.369 0.010 Pb 0.000 0.000 0.000 0.000 0.000 0.000 V 0.239 0.246 0.061 0.033 0.024 0.000 Zn 0.549 0.640 0.089 0.230 0.149 0.012 BR1 DT BV BR1 BV BR1 BV BR1 BV DT BV DT 110414 240414 240414 120414 240414 TF  pH 6.49 7.24 7.2 7.08 6.64 EC 605 512 523 583 590 Ca 139 55 62 74 93 Mg 428 407 429 519 426 Na 561 588 607 604 513 K 123 110 115 78 136 F 72.0 14.6 11.6 −1.0 −1.0 Cl 588.9 605.5 613.7 620.1 570.0 NO2(N) 0.0 0.0 0.0 0.0 0.0 Br 0.0 0.0 0.0 0.0 0.0 NO3(N) 0.0 0.0 0.0 0.0 0.0 PO4 1092.3 642.6 725.8 1142.5 978.1 SO4 2230.6 1765.4 1789.7 2592.8 2054.1 Al 0.013 0.005 0.001 0.008 0.039 Fe 0.015 0.057 0.015 0.015 0.027 Mn 0.618 0.115 0.072 1.440 1.490 As 0.029 <0.006 <0.006 <0.006 0.019 Ba 0.116 <0.001 <0.001 0.005 0.022 Co 0.010 0.004 <0.002 0.010 0.004 Cr <0.006 <0.006 <0.006 <0.006 <0.006 Cu 0.007 0.007 0.006 0.007 0.007 Ni 0.036 0.024 0.024 0.015 0.019 Pb 0.000 0.000 0.000 0.000 0.000 V 0.001 0.000 0.001 0.001 0.001 Zn 0.014 0.011 0.011 0.013 0.011 indicates data missing or illegible when filed