THE CONTROL OF CONTAMINANT DISPERSION FROM MINE WASTE WATER IN THE TROPICS USING DESIGNED WETLAND SYSTEMS

AMD = Acid Mine Drainage

pH = -log[H⁺], where [H⁺] is the concentration of hydrogen ion in moles/L. A low pH(<7) an acidic solution, while a high pH (>7) indicates an alkaline solution.

2. INTRODUCTION
2.1 Climate and Location

Northern Australia lies within the tropics, and as such, experiences a wet-dry seasonal climate (figure 1) (1). In addition, the 'Top End' (defined as the area of Australia north of 15°S) has an annual high temperature distribution (generally between 29 and 34°C in the shade, with temperatures up 77°C being measured in the sun). The monsoonal trough produces annual deluges of rain in the 'wet season', with the highest number of rain days occurring between November and April(2), reaching a maximum average of 21 days in January in Darwin (figures 2 and 3) whilst the minimum falls to an average of near zero for June-August. This region, therefore, has a high rainfall intensity during the wet season. Humidity in the Top End is also high throughout the year with wet-dry variations ranging between 70 and 100% relative humidity. Tropical cyclones are fairly rare events, particularly inland where the majority of the Northern Territory's mines are situated. However, tropical rain depressions are common events both inland and at the coast, and can produce substantial amounts of rainfall.  

Runoff water from mines in the Northern Territory is collected in dams, and only be released under licence (issued under the Water Act, 1992); such regulations are designed to minimise detriment to surrounding aquatic ecosystems. At the Ranger Uranium Mine, where the Uranium Mining (Environmental Control) Act (1978) applies, the Act introduces the concept of a restricted release zone (RRZ), which includes all areas in which material is mined, stockpiled, stored or handled, and the catchment draining rainfall from such areas. All waters in the must be kept within the RRZ, unless release is approved (6).

Generally, dams on mine sites are designed differently according to their purpose. Tailings dams are designed to contain very contaminated material, and so seepage must be minimised. Therefore, the requirements for their construction are very stringent. The performance standards for major dams are set by the Australian National Committee on Large Dams (ANCOLD) ("Guidelines on Risk Assessment 1994") and by specific state legislation. The accumulation of runoff and flood estimation is dealt with in a publication by the Institute of Engineers, Australia ("Australian rainfall and runoff", 1992). 

Freshwater aquatic plants are those plants adapted to growing in/on permanent or semi-permanent water-bodies and have a definite life form related to this environment (11). Species usually found on the edges of water bodies (eg. reeds and rushes) are included, while peat or bog vegetation and brackish water species are not. In this study, we are not interested in saline wetlands and so these will be excluded. Aquatic plants have a strong tendency to accumulate contaminants (by the process known as phytoremediation). A review by Noller et. al. (11) found that manganese is particularly concentrated in freshwater plants (median concentration in plants situated in contaminated sites was 730 μg/g dry weight). Manganese is frequently high in concentration in recessional flows, and is associated with AMD; so plants prove an effective removal mechanism for manganese. They can also concentrate zinc, arsenic, molybdenum, copper, uranium and lead and their ability to accumulate metals far exceeds that of their terrestrial counterparts. Noller et. al. (11) also concluded that the uptake of soluble ions by freshwater plants was inversely correlated with pH and that a low (i.e. acidic) pH increases accumulation of heavy metals by aquatic plants, probably due to the increase in solubility of trace elements under these conditions. Reducing environments also favour the trapping of trace metals in sediments by cation exchange. The presence of organic matter in water/sediment can complex metals and metalloids directly.

Submergent plants were found to be the greatest concentrators of heavy metals. All aquatic plants displayed greater accumulation of heavy metals during their period of rapid growth, which occurs during the wet season in the tropics, and thus corresponds with water release episodes. Also, translocation of heavy metals from the shoots to the roots in the dry season coincides with plant dieback, and hence leads to better entrapment in sediments of heavy metals from decomposing root materials. Studies by Eapaea et. al. (12) at Tom's Gully Gold Mine show that after plant senescence in a wetland, the dried sediments retain arsenic and iron oxide residues.

3. DESIGN OF WETLAND SYSTEMS FOR TREATING MINE WASTE WATER.

Wetlands (both natural and constructed) have been used for effluent purification for several decades (13). Their attraction lies in their cost effectiveness and self sustainability. Until the last ten years wetlands have generally been used in the treatment of municipal waste water, for reduction of phosphorus and nitrogen and to lower biological oxygen demand (BOD). Municipal wastes being of near-neutral pH and having little metal contamination might be considered easier to treat than mine effluent waters. Nevertheless, wetlands are now becoming popular treatments for industrial waste water. There are two main ways in which mining projects can utilise wetland systems. These are: (i) to minimise the dispersion of waterborne contaminants, and (ii) to utilise the fringe zones of existing wetlands to protect the main area of the wetland.

The main sink for metals in constructed wetlands is the substratum, although they are also concentrated by colloidal material (14). The soluble metallic contaminants are the most available to biota. Metals are incorporated into sediments by adsorption, cation exchange, precipitation and chelation. Metal distribution between sediments and the water column is variously affected by the latter's acidity, reducing power, salinity (all of which correlate directly with increased dissolution of trace metals) and organic ligands (which decrease metal solubility) (14). The design of wetlands must therefore take into account all of these factors.

Free metal ions are the main source of toxicity in a wide range of plants and animals (11). Therefore, control of pH is crucial in managing the toxicity of acidic waste waters. Whilst metal uptake by plants is small compared to other removal mechanisms, temporary uptake by plants provides additional capacity for removal. Emergent plants stabilise the substratum by binding it together, and rhizosphere oxygenation aerates the substratum allowing the formation of manganese and iron hydroxides, which scavenge metals and reduce levels of cadmium, copper, nickel, lead and zinc (15). A study of rhizosphere oxygenation by Dunbabin (15) found that metal retention in wetlands was correlated strongly with pH, and only poorly with oxygen concentration and redox potential. However, oxygenation also increases the bioavailability of trace metals to aquatic plants (16). In addition to their other positive features, plants increase microbial and algal activity (which increase metal uptake) (17), provide organic matter on senescence, and act as a filter by slowing water flow, and allowing more of the water column to come into contact with the sediments and deposit particulates and metals (14). Absence of aquatic plants can contribute to erosion of sediments. Floating plants cannot achieve the filtering effect of emergents, and so are not generally used in wetlands for treatment of contaminated waters.

However, a simple wetland system is often not sufficient for the effective removal of contaminants. This is shown clearly, for example, in an experiment where a constructed wetland was used to treat coal mine drainage in Ohio (18). Limestone and mushroom compost were used as a sediment base, and Typha latifolia was planted, as the only wetland plant species. It was found that electrical conductivity, pH, manganese and sulfate were little reduced by the wetland, while iron was reduced by 50 to 60%. Iron hydroxides were precipitating from the water column, thus turbidity increased significantly, and therefore the wetland was simply acting as a sedimentation pond. The difficulty in treating metal-containing waste waters when compared to municipal and sewerage effluent, presents a significant challenge.

It is generally inadvisable to use natural wetlands in treating AMD, because they constitute valuable ecosystems, and have rarely been used successfully (13). Instead constructed wetlands can be used, often in conjunction with other treatments. In many studies, peat, hay, gravel and sawdust have been used as substratum materials in a constructed wetland (19) (peat containing 20-30% organic carbon is available in the Northern Territory from near Bynoe Harbour). Many of these prove to be an expensive substratum for use in a whole wetland but in small sections of a wetland may be viable. Instead, natural sediments are frequently used, with addition of hay at the beginning of wetland operation to provide organic matter for metal entrapment, until plants are properly grown. For AMD careful design of functional zones, including channels containing limestone for acid neutralisation, is required before water enters the wetland since plants and sediments provide little buffering capacity for strongly acidic waters. A riffle, or settling pond is often used before water enters the wetland, to remove iron as an oxide through oxidation of ferrous iron emerging from the dump, otherwise extreme turbidity can result in the system. An anaerobic zone, using peat for example, where water makes no contact with air will remove much of the sulfate, as it is reduced to insoluble metal sulfides (copper, lead, zinc, nickel and cadmium). The wetland stage follows the initial treatment zones and its purpose is to 'polish' the waste water. Sulfate reducing bacteria may also play a part, but are generally difficult to foster, since they are intolerant of acid, and require addition of organic matter

Vigorous plant growth is necessary to ensure success of a wetland. Native plants that are adapted to tropical conditions, should be employed. Moreover, a variety of species is preferred, so that the wetland is less susceptible to attack by viruses or pests. In tropical systems the grass Typha domingensis has proved useful in constructed wetlands, since it is found native in many parts of the Northern Territory's seasonally inundated flood plains (20) and is tolerant of acidic waters with a pH as low as 3. Other acid-tolerant species are Eleocharis spp. and Sclera poaliformis. Bulrushes and Typha spp. are also found in North Queensland. Stockpiling of plants gathered from natural wetlands may be necessary for wetland planting to begin. Aquatic species used in the tropics need to survive the annual cycle through to the end of the dry season and be rejuvenated or grow from seed when the wet season recommences.

3.2 Flow Design

A vertical flow system has been patented by CSIRO, Australia, and this system is more commonly used for hydraulic flow design than horizontal trench style systems, due to possible short-circuiting problems (13). Subsurface flow systems are generally considered more economical in terms of total area required to construct a suitable wetland (21). The optimal flow rates of water entering the wetland are dependent on a number of factors, including the density of plants and the substrate type, but generally, the more concentrated the contaminants in the water, the less the flow rate, to allow maximal contact of water with sediments. Suggested loading rates range from 0.16 to 12 ha/1000m³/day (22). Surface flow systems are generally loaded less than sub surface flow systems.3.3 Classes of Mine Waste WatersMine site runoff waters, which contain solutions from mineral processing residues and waste rock can be classed as follows (23):

1. Mildly alkaline waters
   These waters are derived from groundwater which has contacted carbonate minerals, and are generally bicarbonate-rich. Sources include pit dewatering (24) and seepage from carbonate-containing waste rock material for mildly alkaline runoff. Highly alkaline waters result from cyanide leaching or bauxite dissolution; such waters are not generally released unless highly diluted. Wetlands at tropical locations are able to absorb low concentrations of cyanide and its breakdown products (nitrate/nitrite and ammonia) because low levels of cyanide are a nutrient source (a good example of this is at a gold mine at Masbate in the Philippines where crops are grown on cyanide-rich tailings (25)).
2. Near neutral waters
   The main source of such waters is surface runoff which has ponded and contains little addition of dissolved constituents. Suspended solids may be elevated and may constitute the primary potential pollutant of any released water.
3. Acidic waters
   The sources of AMD have already been discussed. Sulfate is the dominant anion, with iron, aluminium, manganese and magnesium being the major cations present.

3.4 Classes of Wetlands

Wetlands for the remediation of mine waste waters in the tropics can be classed as follows (23):

1. Existing wetlands;
2. Enhanced existing wetlands; and
3. Specific water treatment zones or constructed zones for extreme acid or alkaline conditions, incorporated into artificial wetlands.

The treatment zones in (iii) incorporate special zones which successively modify the waste water to yield improvement in water quality, which the first two classes may not adequately perform on contaminated acidic waters.

The use of wetlands in the tropics for AMD treatment has thus far been limited. Examples of wetlands used in treating mine waste waters of near-neutral pH in northern Australia are:

Ranger Uranium Mine, NT - a constructed wetland which is used to remove uranium, magnesium, manganese and sulfate. Magnesium and some sulfate was not effectively attenuated by the wetland (24)

Tom's Gully Gold Mine, NT - a natural ox-bow billabong wetland which effectively removed a range of transition metals, as well as arsenic from pit dewatering (24)

Woodcutter's Silver-Lead-Zinc Mine, NT - a naturally colonised wetland that removes copper, lead, cadmium and zinc; outflow water is then released into a creek(24)

Hilton base metal mine, Qld - water is degassed to reduce slight acidity, and then a wetland of reeds and algae removes iron, zinc, manganese and thallium (21)

3.5 Fundamental Processes in Wetlands

Optimal performance in constructed wetland systems is achieved by incorporating fundamental processes in the treatment zones in the specific order required to give the desired change in water quality. Summarised, these are as follows:

• the removal of arsenic, uranium, and other heavy metals by passing waters through organic materials;
• the fixation of heavy metals and arsenic using iron oxides;
• acid neutralisation, using limestone, which is consumed and requires replacement following consumption;
• the removal of high concentrations of dissolved iron, which can cause a decrease in the activity of the limestone material; and
• reducing sulfate to sulfide in the presence of the appropriate bacteria in organic-rich material.

The order that the wetland and treatment zones must take is determined by the class of waste water being treated. An example of a plan for a wetland system used to treat AMD, would include the following:

• Treat the iron-rich acid drainage waters emerging from the waste rock dump by allowing it to flow over an open riffle surface of wide surface area. This enables surface oxidation of soluble iron to occur, and thus effective removal of iron from solution.
• Direct 'oxidised' waters into a below-surface organic-rich/limestone zone with air excluded, creating reducing conditions capable of reducing sulfate to sulfide and removing heavy metals as insoluble sulfides. The closer to neutral pH, the more easily the reduction of sulfate is achieved.
• Pass exit waters from the 'reducing' zone to the aquatic plant zone which meanders, in order to give maximum contact of waters with the living plants and associated organic matter. The area of aquatic plants should be as large as possible, allowing sediment filtration and metal removal by sediments, organic matter, microbes and plants.

The above scheme allows zones to be linked in the constructed wetland in a manner that will ensure the continual improvement of waste water quality. Aquatic plants will not, generally, tolerate direct contact with strongly acidic waters. There must be a flow pathway through or around the wetland systems for flood water so that excess waters do not spill over into the wrong places and cause other damage. This consideration applies especially to the tropics since monsoonal storm events can cause large deluges of water. Alternative flow through trenches containing limestone may need to be constructed which will allow for treatment of runoff at high flow rate. Water may need to be pumped in during the dry season to keep the wetland functioning in the following wet season, depending on the plant species used.3.6 Other factorsAn issue of prime importance when constructing wetlands in the tropics is mosquito control. The breeding habits of mosquitoes require them to make use of still water bodies. In the northern parts of Australia mosquito species include: Anopheles spp., which are capable of carrying malaria; Aedes spp. which can carry Ross River virus (which causes epidemic polyarthritis in humans), dengue fever and can infect dogs with heartworm; and Culex spp., which may spread Murray Valley encephalitis virus (26). Some species can disperse many kilometres from their source and bite humans or other animals, and possibly spread the above arboviral viruses. The design of a wetland is critical to its potential for mosquito production; water and vegetation management can be effective for reducing mosquitoes (27). Species of mosquitoes thought to be problematic in wetlands must be considered specifically, but some general techniques in combating mosquito infestation include:

• manipulating water levels;
• selecting appropriate plant species;
• providing water depths unfavourable for mosquito breeding;
• preventing development of stagnant pools of water; and
• utilising animals that feed on mosquitoes and mosquito larvae.

Chemical control is an option not preferred, except as a last resort. In the tropics, in designing a wetland that discourages mosquitoes, it is best to allow for a period in the late dry season when aquatic plants can cease growing and die back, with plant growth recommencing in the wet season, so that watering in the dry is not necessary. Choosing plants that increase diversity of mosquito predators, such as dragonflies and beetles, will help keep mosquito numbers down, and will increase colonisation by insectivorous fish, which will also aid in this respect. If fish are to colonise a wetland, dry-season ponds must become incorporated into wetland design. Ideally, steps should be taken to discourage, if not entirely prevent, birds and other fauna from drinking contaminated water.CONCLUSIONThe tropical wet season results in the accumulation of large amounts of waste water on mine sites. Recharge of waste rock dumps leads to continuous flow of leachates for some months after the end of the wet season.

Wetlands are a common feature of the tropical landscape and may be adapted to treating waste water from mines. For specific control purposes, however, it is usually necessary to design the wetland system taking into account the kind of waste water (whether mildly alkaline, near-neutral and acid) and specialised treatments which may be necessary before waste water is allowed to come into contact with aquatic plants. It is therefore necessary to understand fundamental processes of contaminants in wetland systems, particularly in the case of AMD. Another environmental consideration is the minimisation of the numbers of mosquitoes which may be the vectors of debilitating tropical diseases.

Wetland system development for mine sites in the tropics is still in a state of infancy but looks very promising. Inclusion of features to ensure an even flow of water through the wetland must be incorporated, since, if the water is too deep aquatic plants will not grow.

REFERENCES

1. Butterworth, I., "A review of the Northern Territory wetland rainfall: current and projected", in Wetland research in the wet-dry tropics of Australia workshop, Jabiru Australia, March 1995 pp 212-217, (ed) C M Finlayson.
2. Butterworth, I., "The tropical climate: humidity, storms and lightning", in Proceedings of Power Supply and maintainance of scientific instruments in the tropics, Darwin Australia, November 1991, pp 1-1,2 (eds) J Webb and B N Noller.
3. East, T.J., Uren, C.J., Noller, B.N., Cull, R.F., Curley, P.M., and Unger, C.J., "Erosional stability of rehabilitated uranium mine structures incorporating natural landform characteristics, northern tropical Australia", September 1994, pp 283-298. Z. Geomorph. N. F.
4. Hutchison, Ian P. G. and Ellison, Richard D.(Eds.), "Mine waste Management", Lewis Publishers, 1992, pp 283-298.
5. Gilbert, L. M., "Water flows in mine waste dumps and their implications for AMD control", in the second Australian Acid Mine Drainage workshop, Charters Towers Queensland, 28-31 March 1995, pp 121-126, (Eds.) N J Grundon and L C Bell.
6. Hallenstein, C., and Bastias, J., "Management of Ranger Uranium Mine waters, Kakadu Region, Northern Territory, Australia", in The Third International Mine Water Congress, Melbourne Australia, October 1988, pp 753-761.
7. Salomons, W., "Environmental impact of metals derived from mining activities: processes, predictions, prevention", January 199S, pp 5-23. J. Geochem. Exploration.
8. Zhou, Jiaxin, "Factors Affecting Acid Generation and their Relevance to the Mines of the Northern Territory", in The AusIMM Annual conference, August 1995, pp 479-484.
9. White, Tim, "Stabilising Heavy Metal Waste Underground", June 1995, pp l49-151. Search.
10. Moore, Johnnie N., and Luoma, Samuel N., "Hazardous wastes from large scale metal extraction - A case study", September 1990, pp 1278-1285, Environ. Sci. Technol.
11. Noller, B.N., and Outridge, P.M., "Accumulation of Toxic Trace Elements by Freshwater Vascular plants", Vol 121, 1991 , pp 1-63. Rev. Env. Contamination and Toxicology.
12. Eapaea, Miro P., Parry, David L., and Noller, Barry N., "Retention of Arsenic in a Tropical Wetland Following Dewatering from Tom's Gully Gold Mine", in Proceedings of the 4th Environmental chemistry conference, Darwin NT, 9-14 July 1995, pp ES13-1-ES13-4.
13. Dunbabin, Jane S., and Bowmer, Kathleen H., "Potential use of constructed wetlands for treatment of industrial waste waters containing metals", January 1992, pp 151-168. Sci. Tot. Environ.
14. Petticrew, Ellen L., and Kalff, Jacob, "Water Flow and Clay Retention in Submerged Macrophyte Beds", December 1992, pp 2483-2489. Can. J. Fish. Aquat. Sci.
15. Dunbabin, Janet S., Pokorny, Jan, and Bowmer, Kathleen H., "Rhizosphere oxygenation Typha Domingensis pers. in miniature artificial wetland filters used for metal removal from wastewaters", Vol 29, 1988, pp 303-317. Aquatic Botany.
16. Jackson, Leland J., Kalff, Jacob, and Rasmussen, Joseph B., "Sediment pH and Redox Potential Affect the Bioavailability of Al, Cu, Fe, Mn, and Zn to Rooted Aquatic Macrophytes", January 1993, pp 143-148. Can. J. Fish. Aquat. Sci.
17. Kirschner, Elisabeth M., "Botanical Plants Prove Useful in cleaning Up Industrial Sites", December 1995, pp 22-24. C&EN.
18. Fennessy, Siobhan M., and Mitsch, William J., "Treating coal mine drainage with an artificial wetland", December 1999, pp 1691-1701 . Research Journal WPCF.
19. Steiner, Gerald R., and Freeman Jr., Robert J., "Configuration and Substrate Design Consideration for Constructed Wetlands Wastewater Treatment", in Constructed Wetlands for Wastewater Treatment - Municipal, Industrial and Agricultural", Lewis publishers, 1989, pp 363-404.
20. Finlayson, C. Max, and Von Oertzen, Isabell, "Wetlands of Australia: Northern (Tropical) Australia", in Wetlands of the World I, Kluwer Academic Publishers, 1993, pp 195-304, (eds) D. F. Whigham et al.
21. Jones, D. R., and Chapman, B. M., "Wetlands to Treat AMD - Facts and Fallacies", in the second Australian Acid Mine Drainage workshop, Charters Towers Queensland, 28-31 March 1995 pp 127-146, (Eds.) N J Grundon and L C Bell.
22. Watson, J. T., Rees, S. C., Kadlec, R. H., Knight, R. L., and Whitehouse, A. E., "Performance Expectations and Loading Rates for Constructed Wetlands", in Constructed Wetlands for Wastewater Treatment - Municipal, Industrial and Agricultural, (Ed.) Donald A. Hammer, Lewis Publishers, 1989, pp 319-351.
23. Noller, B. N., "The role of wetlands in controlling contaminant dispersion from mine waste waters", in wetland research in the wet-dry tropics of Australia Workshop, Jabiru, Australia, 22-24 March 1995, pp-161-164 (ed.) C M Finlayson.
24. Noller, B. N., Woods, P. H., and Ross, B. J., "Case studies of wetland filtration of mine waste water in constructed and naturally occurring systems in northern Australia", Vol 29, 1994, pp 257-265. Wat. Sci. Tech.
25. Walker, Simon, "Gold from the mine - crops from the tailings pond", December 1992, pp 18-20, E&MJ.
26. Whelan, Peter, and Kelton, Wendy, "Common Mosquitoes in the Northern Territory - Description of species habitats and Disease potential", Medical Entomology Branch, N.T. Department of Health and Community Services.
27. Russell, Richard C., . and Kuginis, Laura, "Constructed wetlands and mosquitoes - some problems and some solutions," in National Conference on Wetlands for Water Quality Control, James Cook University Townsville, Queenstand, 25-29 September 1995, pp 213-224.

ABOUT THE AUTHORS
Barry Noller (FRACI, PhD University of Tasmania) is president of the RACI and has been Principal Environmental Chemist, Mines Division, Department of Mines and Energy, Darwin, since 1990. From 1980-1990 he was a Research Scientist at the Alligator Rivers Region Research Institute, Jabiru where he undertook substantial work on physico-chemical processes in the tropical environment, particularly the aquatic environment and its interaction with uranium mining. He has been actively involved with the RACI in many areas and was chairman of the Environmental Chemistry Division (1993-1995).

Gretel Parker received her B.Sc. degree with 1st Class Honours in 1995 from the Northern Territory University, majoring in computational chemistry. She has worked in Mines Division, Department of Mines and Energy, Darwin, as a graduate trainee under the supervision of Dr. Barry Noller since June 1995.