Water Management in the Australian Minerals Industry

Transcription

1 34 th IAHR World Congress - Balance and Uncertainty 26 June - 1 July 2011, Brisbane, Australia 33 rd Hydrology & Water Resources Symposium 10 th Hydraulics Conference Water Management in the Australian Minerals Industry D.J. Williams School of Civil Engineering The University of Queensland Brisbane QLD 4072 AUSTRALIA D.Williams@uq.edu.au Abstract: Due to the range of climatic regimes in which they are found, mines may be faced with having too much water, or not enough, but rarely do they have just the right amount of water of the required quality. Water is required for dust suppression during open pit mining and, more particularly, for mineral processing. Water is also liberated in the course of mining and may become contaminated. Dust suppression can be carried out with water of relatively poor quality, such as water already affected by mining activity, provided that runoff does not lead to contamination of the receiving environment. Mineral processing generally requires raw water of relatively good quality, although there are enforced exceptions, such as the need to use hypersaline groundwater in the Kalgoorlie mining region of Western Australia. Mineralised mining and processing wastes can be a source of poor quality water, requiring the control of discharge to the receiving environment. The paper highlights a range of different scenarios, illustrating them with typical examples from a range of climatic regimes and mining situations, focusing on open pit mines in dry regions of Australia. Keywords: Climate, mine wastes, mine water, mineral processing, mining, open pits. 1. INTRODUCTION Mining activities occupy 0.06% of Queensland s land mass, consume less than 5% of the state s water, and employ 1 in 14 Queenslanders, to produce more than 10% of the state s revenue. The minerals industry both uses and makes water, the latter through mine dewatering, although in Australia s dry climate groundwater inflows to open pit and underground mines are generally limited. An exception is the mining of brown coal in the Latrobe Valley, which involves the extraction of about eight/ninths water to one/ninth coal solids, although about four times more water is required for cooling purposes in the associated power stations. The availability of adequate and suitable water to meet mining and processing needs at a given mine site is governed largely by the climatic setting, although in dry regions adequate and suitable groundwater may be available. A key need for water is mineral processing, since it is typically a wet operation. Due to the need for process water of an acceptable quality, it may not be desirable or possible to recycle process water, although in some cases recycling can capture valuable process chemicals. This then results in large volumes of poor quality water that must be managed without harming the receiving environment. In a dry climate, this could involve evaporation, while in a wet climate or where the volume of dirty water cannot be evaporated, water treatment and release to the environment may be necessary. Mine waste storages may generate contaminated seepage and runoff. Waste rock dumps wet up over time due to rainfall infiltration, and may generate contaminated seepage and runoff in the long-term. Tailings storage facilities inevitably generate seepage during their operation due to the large volumes of entrained water discharged with the tailings solids. Seepage rates will reduce on closure of the facility, although ongoing rainfall runoff may pick up surface contaminants, including salinity and acid and metalliferous drainage. Completed open pits may accumulate rainfall runoff and groundwater, remaining a sink in dry climates and a source of potentially contaminated seepage in wet climates, or in dry climates where excess dirty mine runoff is directed to them. Sediment ponds may require regular de-silting during operations and post-closure. ISBN Engineers Australia

2 2. WATER USE IN THE MINERALS INDUSTRY The majority of the world s water is in the oceans and much of the water on and beneath the ground is saline, leaving fresh water of just 1% of the total. It is not only about the need to meet the ever increasing demand, but also about the need to ensure security of supply. This is muddied by the strong advocacy of large water users, politics and state s rights. In his review of water-related issues challenging the sustainability of the minerals industry in Australia, Brown (2003) highlighted that water use in the minerals industry in Australia, the world s driest continent, is only about 2 to 3% of all water used, paling into insignificance beside total agricultural use at 80%, with urban water use at 12%. An estimated 80% of all water used is lost, although this returns to the environment. While water use by the minerals industry may be low, many Australian mines are located in remote, arid areas, making their demand for water a significant proportion of the available supply. This is also true in other dry regions of the world where mining exists, such as the dry south western states of the US, northern Chile, Peru, and Africa, among others. According to Brown (2003), 400 to 1,600 litres (0.4 to 1.6 m 3 or tonne) of water is used per tonne of metalliferous ore processed, and approximately half of this is fresh water. Alumina production uses up to 40 m 3 of fresh water per tonne of bauxite ore processed. Fresh water is required for processing where the processing and/or complexity of the ore chemistry renders the water unusable, while some processing benefits from re-cycling the process water and chemicals, such as cyanide in gold processing. It has been estimated that the Australian minerals industry consumes of the order of 600 Gl/year of water. Groundwater supplies about 20% of Australia s water supply, and is particularly important in arid areas and for remote mines in arid areas. However, Australia s groundwater resource is being depleted, exacerbated by leaking bores, which result in an average 80% water loss. Further, Australia s groundwater is often saline or hypersaline, and the minerals industry has had to adapt their processing to utilise it, e.g. in the Kalgoorlie mining region of Western Australia. The water balance of a mine site is not always as well understood as it should be, and water does not observe mine boundaries. Agricultural users pay perhaps 10 times less for the water they use (currently about $35/Ml, and rising, c.f. about $120/Ml for urban users) than the minerals industry, and produce much less value per unit of water used (agriculture returns less than $500/Ml of water used, c.f. about $40,000/Ml for mining). 3. INFLUENCE OF CLIMATE Figure 1 shows a selection of mine sites covering a range of climatic zones across Australia. Figures 2 and 3 show, respectively, the annual rainfall variability (increasing with decreasing rainfall) and the trends in annual total rainfall across Australia. For the drier parts of Australia, where many of the mines are located, the annual rainfall variability is about an order of magnitude higher than the trend changes in rainfall over the last 100 years. Hence, droughts and floods have a far greater influence on water users than long-term trend changes in rainfall. This is illustrated by the experience at Queensland Alumina during the drought and the subsequent 500 mm rainfall event in February 2003 (Stegink et al., 2003). The drought enforced a mandatory 10% cut in water use in April 2002, which was readily achieved by the re-use of some wastewater and improved process control. A further 15% cut was mandated in November 2002, which was primarily achieved by the re-use of secondary-treated Gladstone effluent, plus the re-use of Queensland Alumina s treated effluent and increased % solids of the tailings. A further cut to 50% would have required desalination or salt water cooling to replace evaporative (water) cooling of the plant and power station, but this was not enforced due to the drought-relieving rainfall. The water use savings of 25% were retained. Other maps provided by the Australian Bureau of Meteorology (ABoM), include the extent of Australian river basins, average annual rainfall, the average annual number of rainfall days (increasing with increasing rainfall), average annual sunshine hours/day (high, resulting in high pan evaporation), average annual solar exposure/day (high, resulting in high pan evaporation), average annual daily minimum temperature, average annual daily maximum temperature, average annual pan evaporation (high and generally much greater than rainfall), and average annual evapotranspiration (generally approaching the available rainfall). 2202

3 Rum Jungle Argyle Century Kidston Mt Whaleback Granny Smith Nickel West Kalgoorlie St Ives Mary Kathleen Olympic Dam Mt Morgan Cadia Savage River Figure 1 Selection of mine sites covering a range of climatic zones across Australia (after Australian Bureau of Meteorology, Rum Jungle Rum Jungle Argyle Century Kidston Argyle Century Kidston Mt Whaleback Mary Kathleen Mt Whaleback Mary Kathleen Granny Smith Nickel West Kalgoorlie St Ives Olympic Dam Mt Morgan Cadia Granny Smith Nickel West Kalgoorlie St Ives Olympic Dam Mt Morgan Cadia Savage River Savage River Figure 2 Annual rainfall variability across Australia (after Australian Bureau of Meteorology, Figure 3 Trend in annual total rainfall across Australia (after Australian Bureau of Meteorology, Based on the ABoM maps, the threshold average annual rainfall for net infiltration is about 250 mm, corresponding to Kalgoorlie, as shown by the plot of annual rainfall / actual evapotranspiration versus annual rainfall in Figure 4. Water discharged with tailings and, to a lesser extent the wetting up of a waste rock dump, adds to the rainfall, increasing the moisture excess. 4. SEEPAGE AND RUNOFF FROM WASTE STORAGES The minerals industry is moving to minimise water use, maximise water recovery, and minimise contaminant transport to the environment, e.g. by recycling process water where possible, tailings thickening and paste, and moving towards dry processing. Mine closure and decommissioning raise further important water management issues, including the ongoing control of contamination and sediment, and mine water discharges. In the following sections, the seepage and runoff of potentially contaminated water from surface waste rock dumps (WRDs) and tailings storage facilities (TSFs) are discussed. 2203

4 RAINFALL / EVAPOTRANSPIRATION Kalgoorlie, WA, average annual rainfall = 250 mm Net evapotranspirative Net infiltrative RAINFALL (mm/year) Moisture excess Figure 4 Rainfall / evapotranspiration versus rainfall (after Williams and Williams, 2007) 4.1. Waste Rock Dump Water Management In Australia, waste rock dumps (WRDs) are generally only a minor component of an open pit mine site s water balance. However, if the waste rock is reactive, the loose-dumped construction of surface WRDs leads to the waste rock being well-oxygenated, causing the reactive waste rock to oxidise. The oxidation products are then available to be transported as the dump wets up from rainfall infiltration. Where contaminated seepage emerges from the dump, there is a need to limit the net percolation of rainfall and/or the air permeability through the dump, using an effective cover system Rainfall Infiltration, Storage and Base Seepage In the semi-arid or arid climates in which many of Australia s mines operate, waste rock emerges from an open pit with a low moisture content. From the outset, a surface WRD closes off evaporation, while allowing rainfall infiltration, acting like a sponge. Much of the rainfall infiltration will initially go into storage within the voids in the dump, with any excess ultimately emerging as seepage at the toe and into the foundation. The wetting up of the dump will progress along preferred pathways as the ability of the waste rock pores to store water is exceeded, this occurring well below the fully saturated state since the waste rock will achieve a sufficiently high hydraulic conductivity to pass further rainfall infiltration. At this stage, the amount of rainfall infiltrating the dump is matched by the amount of seepage emerging from its base, and continuum breakthrough has been reached. The base seepage will go into the foundation, if it is sufficiently permeable, and/or collect in buried surface channels and emerge at low points around the toe of the dump. The time taken to reach continuum breakthrough is a function of the average annual rainfall of the site, the height of the dump and the composition of the dump. Coarse-grained, durable, fresh waste rock only needs to reach about 20% saturation of its voids to be free-draining, while fine-grained, wellgraded, weathered waste rock needs to reach about 60% saturation of it s voids to be free-draining. The higher the average annual rainfall, the higher the dump and the finer-grained the waste rock, the longer it takes to reach continuum breakthrough, as illustrated in Figure 5 (Williams and Rohde, 2007). For a typical dump height of 30 m and a typical mine life of 20 years, during which the WRD is left uncovered, a semi-arid to arid climate would cause continuum breakthrough in about 25 years. Once continuum breakthrough has been reached, any oxidation products on the surfaces of the waste rock particles can be transported with the seepage. Seepage increases exponentially towards continuum breakthrough, and then responds to individual rainfall events. Whatever time it takes to reach continuum breakthrough, it will take a similar length of time for the WRD to drain down if it were 2204

5 covered with an effective low net percolation cover, and it would drain down at a rate diminishing exponentially with time towards whatever net percolation rate is achieved by the cover. The amount of rainfall infiltration into the dump can be restricted by sloping the traffic-compacted top surface to avoid ponding, although discharging the top runoff over the steep sides of the dump will likely induce erosion. The loose-dumped outer slopes of the waste rock dump should be constructed of benign waste rock of sufficient thickness to produce clean runoff and seepage. AVERAGE RAINFALL (mm/year) Start for 15 m height Full for 15 m height Start for 30 m height Full for 30 m height Start for 60 m height Full for 60 m height Start for 120 m height Full for 120 m height Typical mine life of 20 years Typical dump height of 30 m CONTINUUM BREAKTHROUGH (years) Figure 5 Estimated time for continuum breakthrough of weathered waste rock dumps Williams and Rohde (2008) collected, from the literature and other sources, infiltration and seepage data (expressed as a % of annual average rainfall) for uncovered WRDs comprising weathered waste rock, to produce Figure 6. By plotting the infiltration and seepage against WRD thickness / its age, the height of the WRD and the number of years it has been exposed to rainfall are captured. Low level and old uncovered WRDs tend to infiltrate about 75% of the incident rainfall, while typical WRD heights infiltrate about 50% of the incident rainfall during their operational lives, partly due to runoff being inhibited by the perimeter windrows constructed for plant safety purposes. Low level and old uncovered WRDs tend to seep about half of the incident rainfall, while typical WRD heights infiltrate only a few percent of the incident rainfall during their operational lives, since they typically remain well short of continuum breakthrough over this timeframe. The remainder of the infiltration goes into storage within the WRD. The proportion of rainfall that does not infiltrate goes to runoff and evaporation. INFILTRATION or SEEPAGE (% of average annual rainfall) Infiltration Seepage Log. (Infiltration) Log. (Seepage) WASTE ROCK DUMP HEIGHT / AGE (m/year) Figure 6 Infiltration and seepage versus weathered waste rock dump height / age (after Williams and Rohde, 2008) 2205

6 WRDs Post-Closure If WRDs are left uncovered on mine completion, they will continue to infiltrate rainfall, at a slowly diminishing rate as the surface naturally forms a hardpan of reduced hydraulic conductivity, and they will eventually reach continuum breakthrough (if they hadn t already reached this state during the life of the mine). If the waste rock is inert, the ongoing wetting up and eventual continuum breakthrough of the dump will be of no great concern. However, if the waste rock is reactive and likely to generate salinity and/or acid and metalliferous drainage, attempts should be made to limit net percolation through the construction of a low net percolation cover. The two most common cover types for dry climates are store and release covers and rainfall-shedding covers, of which the former are more robust and can be effective if properly designed and constructed (Wilson et al., 2003). The critical elements of a store and release cover are a sealing layer at the base of the cover to hold-up rainfall that infiltrates the upper rocky soil mulch growth medium, and the appropriate choice of sustainable vegetation species. Rainfall-shedding covers suffer from erosion loss, and the difficulty of sustaining a vegetative cover in a seasonally dry climate Tailings Storage Facility Water Management Tailings storage facilities (TSFs) are perhaps the key determinant of the mine water balance during their operation, when large volumes of tailings water, many times larger than the low rainfall experienced at many of the mine sites in Australia, are discharged to the TSF. While considerable water remains entrained within the tailings, the equally considerable supernatant water may or may not be suitable to be re-used in the processing plant, and water is lost to evaporation and seepage. On closure, the TSF is subject to very much lower ongoing rainfall inputs. With transient ponding occurring on the surface, evaporation and seepage reduce towards background levels. Where there is a potential for contaminated seepage from a closed TSF, it is desirable to drain down the tailings to limit seepage volumes, and this may require the construction of a spillway to remove rainfall runoff from the top of the TSF Unlined TSFs As detailed by Williams and Williams (2008), the discharge of tailings at an initial solids concentration of 25 to 50% introduces the equivalent of about 1,000 to 750 mm/year of water, compared with average annual rainfall of 200 to 400 mm/year over much of Australia s mining regions, giving a total water input to the TSF in the range from 1,400 to 950 mm/year (about 6 to 3 times average annual rainfall). On deposition, the tailings slurry will sediment and consolidate, releasing supernatant water to form a surface pond, encouraging seepage into the foundation and containment walls. Some of the water discharged with the tailings remains permanently entrained within the tailings. Assuming a typical average rate of rise of tailings of 1 m/year and an average final dry density for metalliferous tailings of 1.5 t/m 3, perhaps 350 to 250 mm of water will remain entrained within the depth of tailings deposited each year, and 550 to 250 mm of water will be available for recycling, with increasing initial % solids from 25 to 50%. The remainder of the water will be lost to evaporation and seepage. Where the supernatant water is unsuitable for re-use in the plant, it is generally stored either on the TSF or, preferably, in separate evaporation ponds, with the ponded water presenting a potential hazard to birds and increased risk of significant groundwater impacts. The amount of seepage emerging from a TSF can be restricted by placing the tailings at as high a % solids as possible, and efficiently removing supernatant water. Tailings deposition can be cycled between cells to maintain unsaturated conditions within the foundation to the TSF and limit seepage to the foundation; however, seepage losses will still be of the order of 10 times the natural groundwater recharge rate in dry regions of 1 to 50 mm/year Lined TSFs Liners have the potential to greatly reduce the seepage from TSFs, both during their operational phase when water inputs are high and post-closure when driving heads reduce. If the liner is effective, the reduction in seepage losses will mean an increase in the volume of supernatant water of 2206

7 up to 35% compared with having no liner. If this increased supernatant water is not removed from the TSF, there will be a reduction in the desiccation and shrinkage of the tailings, with consequences for the storage capacity of the TSF, the acceptable rate of rise, and the practicability of upstream construction on top of the tailings. If the increased supernatant water is not suitable for re-use in the plant, it may require the development of expensive separate evaporation ponds, which will also have to be effectively lined. The effectiveness of liners depends on their integrity, which in turn depends on the materials used and the construction and maintenance methods employed. Their effectiveness also depends on the hydraulic gradient they are subjected to, which in turn depends on the presence or absence of underdrainage, the applied hydraulic head, and the thickness of the liner. Liners can comprise natural earthen materials, geomembranes (e.g. high density polyethylene), and composites such as geosynthetic clay liners (GCLs). For TSFs, there is a compelling argument for underdrainage in association with liners to limit the hydraulic head on the liner. It should be recognised that underdrains become less effective over time as the tailings consolidate around them, block them, and if they become affected by the changing chemistry of the tailings and tailings water. For evaporation ponds, great care must be exercised in designing and constructing an effective liner to accommodate the high hydraulic gradient applied Tailings Thickening The capital costs of thickened and paste tailings production and disposal are a major impediment to their use, despite the potential operational and long-term benefits. Operational benefits include the greater in-plant recovery of process water and process chemicals, reduced seepage, and better use of available storage volumes. Long-term benefits include smaller tailings storage footprints, and reduced seepage to the environment TSFs Post-Closure Once tailings deposition ceases, the huge volume of water discharged with the tailings is taken out of the equation, making the TSF much less of an environmental issue. However, the water entrained within the tailings during the operation of the TSF will continue to seep for many years. Seepage and the transport of contaminants into the foundation, and possibly also through the containment wall, will diminish approximately exponentially with time as the tailings drain down, consolidate and desiccate, causing reduced hydraulic conductivity. In Australia s semi-arid mining regions, which naturally display very low recharge rates to groundwater, and the very low hydraulic conductivity of the desiccated tailings surface, a cover to limit infiltration may not be necessary, provided the tailings are not allowed to fully re-saturate due to long-term concentrated ponding of rainfall runoff. However, a cover may be desirable and necessary to limit dusting and/or for revegetation purposes. Overtopping of the outer slope of the tailings containment should be avoided to limit erosion. A spillway may be required to prevent the possibility of overtopping of the TSF and ensure that rainfall runoff is not able to concentrate and pond for long periods of time on the tailings surface. 5. OPEN PITS Open pits will typically fill with water to some extent after the end of mining. In Australia s dry climatic regions, this is mainly driven by rainfall runoff rather than by groundwater inflows, and may take 10 to 30 years, or longer. The chemistry of the pit water will be governed by the geochemistry of the pit walls and any mine wastes stored in the pit, the quality of the water directed to the pit and its dilution by rainfall runoff, and the length of time it takes for the pit to fill with water. Where excess rainfall runoff is directed to the pit, the pit water level may exceed that of the surrounding groundwater and contaminated pit water will become a source to the environment. If the pit receives just incident rainfall, its water level is likely to remain below the level of the surrounding groundwater and it will remain a sink and not flow to the environment. 2207

8 6. CONCLUDING REMARKS The Australian minerals industry accounts for 2 to 3% of the nation s water usage, and is a reasonably efficient user. However, given the remoteness and aridity of many Australian mine sites, their demand for water may be a significant proportion of the available supply. The return from mining on a water usage basis is very high. Of key importance to the minerals industry is maintaining security of water supply, while preserving the competing water entitlements of other stakeholders and recognising the limited water resource. There is a need to understand the true cost and true value of water, based on the potential loss of production. There is also a need to move from a reactive and ad-hoc water management to holistic and proactive water management. Rain leads to apathy; drought leads to awareness, concern and panic; and these are relieved by subsequent rain. The Australian minerals industry does some things well. Mining operations have been able to limit water use even in the face of increased production, and where under water supply pressure, they have been able to reduce water consumption to as little as one third. Mining companies are good at negotiating water licenses for new mining projects, and accommodating water restrictions when these are forced by drought. Best practice water management is emerging, high solids tailings disposal is being implemented, and dry processing is being researched. Tangible incentives need to be provided by Governments and Mining Companies for reducing water usage. Industry best practice water management needs to be defined to set target water usage rates, which could be used as the basis for imposing water restrictions during times of drought. More stakeholder partnerships between mining operations and local stakeholders should be developed, and the minerals industry needs to sell good water management practice more effectively. 7. ACKNOWLEDGMENTS This paper draws on expertise and experience collected by the author over the last 20 or more years from research carried out at the Golder Geomechanics Centre within the School of Civil Engineering at The University of Queensland, and his engagement with the minerals industry. The contributions of the many co-workers and colleagues are gratefully acknowledged. 8. REFERENCES Brown, E.T. (2003). Water for a sustainable minerals industry A review. Proceedings of Water in Mining 2003, Brisbane, Australia, October 2003, pp Stegink, H.D.J., Lane, J., Barker, D.J. and Pei, B. (2003). Water usage reduction at Queensland Alumina. Proceedings of Water in Mining 2003, Brisbane, Australia, October 2003, pp Williams, D.J. and Rohde, T.K. (2007). Strategies for reducing seepage from surface waste rock piles during operation and post-closure. Proceedings of Second International Seminar on Mine Closure, Santiago, Chile, October 2007, pp Williams, D.J. and Rohde, T.K. (2008). Rainfall infiltration into and seepage from rock dumps A review. Proceedings of First International Seminar on the Management of Rock Dumps, Stockpiles and Heap Leach Pads, Perth, Australia, 5-6 March 2008, pp Williams, D.J. and Williams, D.A. (2008). Possible impacts on mine water balance arising from lining of a tailings storage facility in Western Australia. Proceedings of Tailings Management for Decision Makers, Perth, Australia, 3-4 December 2007, 5, pp Wilson, G.W., Williams, D.J. and Rykaart, E.M. (2003). The integrity of cover systems An update. Proceedings of Sixth International Conference on Acid Rock Drainage, Cairns, Australia, July 2003, pp