Mining, Mineral Processing & Extractive Metallurgy

Paul Mitchell & Alyson Warhurst

INTRODUCTION

Mining and the Environment - Past and Present

The availability of rich geological resources in tandem with market conditions largely outside the control of the metal-producing enterprises have resulted in a sector characterised by a low level of technological innovation and poor environmental performance. Corporate strategies have traditionally focused on the discovery and acquisition of high grade and easily accessible mineral deposits, and on increased scale of production to offset declining ore grades.



Although permanent, temporary or transient environmental impacts are common, their association with mining and milling operations is not inevitable. For example, of thirty-three mining and milling operations surveyed in 1989-90 by Environment Canada, approximately half had no adverse environmental impacts, while a further 15% had only minor effects (Eaton et al, 1994). However, as is the case for many heavy, process-orientated sectors, mining has had limited success in altering the widely-held perception that it must be an intrinsically "dirty" and polluting activity. Despite pro-active initiatives in the fields of waste management, pollution prevention and multi-party dialogue with stakeholders, societal perception of the general industry continues to revolve, to a large degree, around consideration of "sins of the past" rather than current and state-of-the-art operations.

---

In the past, health and safety regulation has been a major driver behind technological change alongside the need to improve process energy efficiency. However, the emphasis has now altered: after a period of limited technological change a spur to technology development in the minerals industry has been applied through (a) public concern regarding adverse environmental effects and (b) the development and implementation of environmental regulation that obliges firms to mitigate or prevent such effects. However, there are significant constraints to the performance of operators in terms of process emissions, and these have substantial implications for the development of "zerowaste" operations and the creation of environmental and social performance indicators measured against this ideal.



Dynamic companies have responded to societal perceptions and other drivers by improving environmental performance as part of the renewed drive for competitive edge. In many cases environmental performance and standards continue to improve. Leading companies are now beginning to operate within the regulatory standards and guidelines of their home country at overseas sites where less stringent legislation or enforcement would otherwise apply. This shows that commitment to minimising environmental impacts is growing and that competitiveness and enhanced environmental performance are not mutually exclusive, although they may be constrained by factors outside the immediate control of the company. Similarly, it is becoming increasingly clear that technological incompetence is not compatible with the concept of sustainable development and that mining, mineral processing and extractive metallurgical technologies all play a significant role in determining the environmental performance of an operation and its contribution to sustainable regional and national development.



This paper examines the phases, operations, activities and processes that make up the "mining life cycle", associated process and related emissions and the potential for geological and mineralogical factors to influence the site- and medium-specific "baseline" environmental performance of operators. It then goes on to explore technological and waste management trade-offs that may shift upwards or downwards the "baseline" in one or more of the environmental media - soil, air and water and how the mining industry is seeking to enhance its performance and mitigate its environmental impacts. The paper focuses particularly on base and precious metal operations, where some of the most complex and intransigent issues occur. However, other commodities (e.g. coal, industrial minerals) are also considered where appropriate.

Definition of Terms

In dealing with mining and environment issues, it is important that key concepts are properly defined. In the past, this has not been the case, and consequently, confusion has arisen, particularly amongst those working outside the industry. Definitions that are considered by the authors to be essential to the understanding of issues relating to technology and emission include:

• Contaminant (or contamination)


• Pollutant (or pollution)


• Hazard


• Risk


• Process release (or environmental releases)


• Environmental effect


• Environmental impact (or environmental damage)

The term contaminant (or contamination) tends to be used interchangeably with pollutant (or pollution). However, the two are distinctly different. The former is defined as "a substance found in a given medium at a concentration higher than that which one would expect from other considerations and where the source of the additional concentration of the substance appears to be human activity" (Davies, 1992). A contaminant is only a pollutant if the observed concentration is high enough to cause harm to some organism. Proving that pollution, rather than contamination, has occurred is often a difficult task, particularly retrospectively (Mitchell, 1994).

Hazards and risks are often confused, particularly by those working outside the industry. Hazards are a physical situation with a potential for human injury, damage to property, damage to the environment or some combination of these, while a risk is the likelihood of specified undesired events (with specific consequences) occurring within a specified period in specified circumstances arising from the realisation of a specified hazard. For further information on hazards and risks, and their significance for the development of ESPIs, see "Issues Paper" by Thomas Coleman.



The terms process release (also known as environmental release), environmental effect, environmental impact (or environmental damage) are often used interchangeably. However, this abuse of terminology promotes the erroneous belief that a release will inevitable cause an impact, and that cause-and-effect can always be clearly identified. The concepts should be seen instead as a logical sequence, which begins with the process release, and may end with environmental damage, depending on other mitigating factors such as dilution, biodegradation, attenuating mechanisms or resilience of the receiving ecosystem (see "Issues Paper" by Kevin Franklin).



Process (or environmental) releases can be defined as a transfer of material or energy to the external environment (i.e. everything outside the process). Examples include:

• The transfer of a solid waste or liquid effluent outside of the mill building.


• The discharge of gaseous emissions from a stack.


• Transfer of a metal from the solid mineral phase to water in underground and open pit workings.


• Dust from an open pit.


• Noise from blasting.

In most cases, potential environmental effects are linked to the material being transferred. However, some may be related to the consequences of making that transfer (e.g. subsidence in underground mining or slope stability in open pits relates to the removal of ore and overburden).



Environmental effects can be defined as a measurable change in the external physical environment (i.e. a measurable disturbance in the existing system) that results from an environmental release. Therefore, certain environmental releases may cause no measurable effect due to mitigating processes in the external environment. Effects, while measurable, may have no discernible consequence for ecosystem and/or human health.

Environmental impact (or damage) implies that the effect of an environmental release is such that mitigating processes are not sufficient to prevent degradation of the external environment. In these cases either the external environment can not return to its previous state without human intervention or the release has a direct and measurable consequence for ecosystem and/or human health.



To facilitate the development of credible environmental performance indicators and to promote a high standard of environmental performance, a broad and deep understanding of the causes and nature of potential releases, effects and impacts is essential. This understanding should underpin and inform decisions throughout the life cycle of an operation.



It should be noted that environmental releases may be localised and contained within the site boundary. Impacts may be transient (often acute, associated with spills or accidental discharges), temporary (related to operational discharges, ceasing when operations are terminated) and chronic (long-term, often those associated with ore extraction and waste disposal). Typical transfer mechanisms are wind and water erosion of contaminated solids or dissolution of contaminants into the aqueous phase. The movement of intermediate products off-site for further processing (e.g. the transport of concentrates to a smelter that may be hundreds or thousands of miles away) is another mechanism by which the spatial and temporal environmental impact of an operation can be extended. Therefore, any analysis of the environmental burden arising from a particular operation must address the entire life cycle, from run-of-mine ore through to the final product, including any and all off-site operations. This may lead to the choice of a technology that would not otherwise have been considered appropriate or necessary.

The Roots of Waste Production

Metals and other mineral resources are rarely in found in a sufficiently pure state to be sold in an "asmined" form. Metals are often found in chemical combination with oxygen (as oxides), sulphur (as sulphides) or other elements (e.g. chlorides, carbonates, arsenates, phosphates, etc) and physically mixed with less valuable or valueless minerals phases (e.g. silicates). Non-metal mineral resources (e.g. coal, industrial minerals) also normally contain physically- or chemically-entrained impurities in their undisturbed state. Valueless or sub-economic minerals associated with the target or economic mineral(s) are generally known as gangue.



Inputs from mining (extraction) to mineral processing/extractive metallurgy invariably contain varying levels of gangue and low-value minerals which exit the process or processes in a number of forms (e.g. as contaminants in the primary product, as separate saleable by-products, solid wastes, dissolved species, gases, fumes, suspended particulates). During the extraction and processing of the ore, mass is conserved (i.e. mass inputs equal mass outputs). However, the physical and chemical

characteristics of the input phases may be modified by interactions with process chemicals and/or the process itself. For example, depending on the point at which it is rejected from the process, gangue may be disposed of in an as-mined state (e.g. waste rock), as tailings (e.g. following mineral processing), as slags (e.g. after smelting) or as other waste products (e.g. dusts, sludges from water treatment, spent ore from leaching etc). These various wastes may also contain significant quantities of the target mineral or metal due to inefficient processing, technological limitations or mineralogical factors. Deliberate (engineered) modification of the physical and chemical characteristics of waste outputs represents a potential tool for reducing the environmental burden of waste disposal, and this is discussed in more detail below.



In essence, the transfer of wastes from the process to the external environment is a consequence of dealing with "impure" feed materials and the "reversal" (in relatively short periods of time) of physicochemical mixing of valuable and non-valuable mineral phases by geological and geochemical processes that has taken place over millions of years. In essence, the industry processes low concentration raw materials into higher concentration products with significant energy input (Stewart and Petrie, 1998). Feed materials are also thermodynamically stable and therefore to impose increased order on the system (by decreasing the disorder represented by the dispersion of the metals within the ore) a significant energy input is required (Stewart and Petrie, 1998). Until there is a paradigm shift in mining methods that allows metals or mineral phases to be selectively recovered with little or no gangue, waste generation and transfer to land, water or air can be taken as a given. However, the manner of that transfer and the form of the waste that is transferred is more directly under the control of the operator, and it is this that forms the focus of this paper: the potential for, and constraints to mitigation of geological, mineralogical and geochemical factors through technology choice and management. Technology choice and management must also be considered in the context of the "requirements" for an appropriate level of protection for the receiving environment (see "Issues Paper" by Kevin Franklin).



In non-ferrous metal mining, gangue is normally the major component of an ore body. Nowhere is this more apparent that in the case of gold where the concentration of valuable material is so low (e.g. normally 5 g t-1 or less) that effectively all of the mined ore is disposed of as waste (unless other valuable components such as base metals are also present). Other mineral resources may have less gangue relative to the target mineral, but disposal of gangue-related wastes normally remains a significant issue. "Average" figures (based upon a survey of Canadian metal mines) indicate that 42%

of the total mined material is rejected as waste rock, a further 52% from the mill as tailings, an additional 4% from the smelter as slag, leaving a valuable component of just 2% of the originally mined tonnage (Boldt, 1967; Godin, 1991). In effect, such sites are as much about waste disposal as they are about resource extraction. To place this in a wider context, Table 1 shows ore and waste production figures for the USA in 1991 for a range of mineral resources.



Table 1 Estimated ore production, average grade and waste generation, USA 1991

(Worldwatch Institute, 1992; Rogich 1992)



* Waste figures do not include overburden



Clearly, the higher the concentration of valuable components, the lower the tonnage of waste produced per tonne of ore extracted. However, there is also a trade-off between grade (the concentration of metal(s) in the final or intermediate product, e.g. concentrate) and the recovery (percentage of the total valuable metal(s) present contained within the feed to the processing plant reporting to saleable products). It is possible to maintain very high grade by rejecting a significant fraction of the input material (i.e. by accepting low recovery) or very high recovery by excessively diluting concentrates with lower grade material. However, neither of these two extremes is normally the optimum economic solution. In simple terms, this is gauged by comparing the revenues generated by additional recovery of metal(s) against the capital and operating costs of doing so, within the greater context of technical feasibility. Without exception, some part of the target metal(s) will report to the wastes, along with the gangue minerals.



Waste materials are unavoidable, both during mineral processing and later extractive metallurgical stages. Thus adopting a waste minimisation approach to waste management problems within the minerals industry will not of itself present a comprehensive solution(Stewart and Petrie, 1998). Wastes are usually very diverse in nature, with some characteristics being set by technology choice and operational management, while others relate directly or indirectly to the geological, mineralogical or structural characteristics of the orebody.



Environmental liability arising from the disposal of wastes is difficult to properly quantify due to the variable dispersion pathways in different receiving environments, and the ultimate fate of the wastes and the ore- and process-related chemicals that they contain. The issues paper by Kevin Franklin addresses biodiversity and the "assimilative capacity" of the receiving environment and notes the significance of the rate of discharge and mobilisation in determining the reversible or permanent damage caused by waste releases.



It is apparent that the process of waste disposal related to mining activity is a significant source of potentially harmful elements in the natural environment.



Thornton (1995) adapted figures from Nriagu and Pacyna (1988) to estimate the anthropogenic input of a range of metals to aquatic ecosystems from base metal mining and processing (Table 2) and disposal of mine tailings on land (Table 3).

Table 2 Anthropogenic inputs to aquatic ecosystems from base metal mining and dressing(10⁶ kg year ^-1)

However, inputs do not necessarily result in damage to the environment, many mitigating factors may exist, some of which relate to the process (e.g. the chemical and physical characteristics of the wastes) and other which relate to the external environment (e.g. climate, topography, ecosystem characteristics).



However, any unit operation within the life cycle of a mining operation has the potential to produce an environmental effect or impact. Typically the potential arises from the deliberate (regulated) and accidental (non-regulated) discharge of solid, liquid and gaseous waste products. The characteristics of the discharges, the nature of the receiving environment and the distance over which the discharges are transported are major factors in determining the magnitude of the effect or impact. Societal values and preferences also play a significant role in determining how certain discharges are viewed by various stakeholder groups: this more subjective adjunct to the quantifiable and measurable discharge and receiving environment characteristics therefore sets, in part, the site-specific environmental "footprint" of an operation.

THE MINING LIFE CYCLE, PROCESS RELEASES AND RELATED ISSUES

Mining

Introduction

Mining is the first operation in the commercial exploitation of a mineral resource. It can be defined as the act of tunnelling and digging out of the ground in order to recover one or more component parts of the mined material (Dunster and Dunster, 1996).



In broad terms, there are three types of mining: surface, underground and in-situ (solution mining). The latter is somewhat limited in its application, although it is sometimes used to exploit residual mineralisation as grades drop at surface or underground mines. Surface mining is dominated by open pit (e.g. base and precious metal ore extraction) or open cast (e.g. coal operations) methods. Surface and underground mining usually occur independently of one another (e.g. the choice of method is "either/or") although open pit mining does occasionally occur in areas already partly worked by underground methods. Similarly, underground methods are sometimes used to extract ore from beneath, or in the vicinity of, pits where further extension of the pit itself is not economic or technically feasible.



Irrespective of the method employed, mining is always accompanied by processing of some description. For relatively pure or homogeneous materials processing may be limited to crushing and sizing (e.g. some natural zeolite extraction, quarried rock) or washing (e.g. some coal operations). Simple processing such as this is only possible where the target mineral forms the majority of the material mined. In these cases, the main environmental releases, effects and impacts are associated primarily with the mining itself rather than subsequent processing.



Factors that influence the choice of mining method include the size, shape, dip, continuity, depth, and grade of the ore body; topography; tonnage; ore reserves; and geographic location.

Surface mining

Surface mining can be divided into three sub-groups:

• Open pit or bench mining - deposits that are deep but have restricted width.


• Open cast or strip mining of relatively shallow deposits.


• Alluvial mining of unconsolidated near surface deposits on land or beneath water.

Compared to underground mining, surface mining methods are generally considered to allow for a higher degree of worker safety, greater flexibility in extraction (the capacity to practice selective mining and grade control) and lower development and maintenance costs (due to the requirement for fewer specialised systems). However, the major benefit lies in economy of scale, as large capacity earth moving equipment can be used to generate high productivity.

Open pit mining

Open pit mining is normally used for steeply dipping beds or veins or for massive irregular deposits. In open pit mining, overburden (non-mineralised soil/rock that covers an ore body) and waste rock (poorly mineralised or very low-grade soil and rock that are within the ore body or surrounding it) are first removed from the entire area of the final planned pit, often in very large quantities.3 Sub-economic ore may also have to be removed to expose the economic reserve. The ore itself is then removed for further processing.



The mine shape is formed by a series of benches or terraces arranged in a spiral or in levels with interconnecting ramps. Open-pit mines may reach several thousand feet below the surface. The pit is deepened in a sequential manner using benches that also serve as haulage roads for the removal of ore from the pit. Restoration at the end of the operation can be very expensive particularly if it involves backfilling, as the wastes are often dumped at some distance to avoid obstructing the removal of economic ore.



Overburden and waste rock are often used during the operation and closure of a mine (e.g. inert waste rock for bunding, soils for reclamation at closure). However, if they are contaminated with significant concentrations of potentially harmful minerals (albeit at sub-economic concentrations), they must be disposed of as wastes.

Open cast mining

In open cast mining, extraction proceeds laterally rather than vertically and this method is therefore most suitable for shallow deposits. A "strip" of overburden is first removed to expose the underlying deposit. The deposit is then worked-out, following which another strip is prepared, with the overburden being placed preceding (worked-out) strip. Thereafter, restoration is done in a rolling fashion, with only two "active" strips at any one time (one being stripped, one being backfilled).

Alluvial mining

Alluvial mining is the cheapest method and when applied on a large-scale can be used to mine ores of very low contained value (ie: low grade, low metal value or both). For example, in South East Asia tin ores containing as little as 0.01% tin are mined by alluvial methods although the ores have a contained value of less than £1/t (Wills, 1997). Alluvial mining does not involve drilling, stripping or blasting, but rather the recovery of the deposit through physical means such as scraping, high pressure water jets or suction (for under-water deposits).



As alluvial deposits often contain liberated or partly liberated target mineral(s) (i.e. they occur in particles physically separated from the gangue minerals), crushing and/or grinding is often unnecessary, reducing the cost of processing dramatically (see Mineral Processing, below).

Underground mining

Underground mining methods are usually employed to mine richer, deeper, and smaller ore bodies where open pit methods would be impractical. Underground mining operations are a complex combination of tunnelling, rock support, ventilation, electrical systems, water control, and hoists for the transportation of people, ore, and materials. Consequently, underground mining is less flexible than surface working and deviations in the production rate are more difficult to accommodate. Because mining may be proceeding in several different underground locations to ensure an appropriate level of production, forward planning and control of mine development are extremely important.



Mines in shallow dipping ore deposits take the form of flat or shallow dipping parallel-sided cavities. In room and pillar mining, pillars of unworked ore are used to support the roof. The size of the pillars is dependent on the strength of the ore being exploited. Areas that have been worked-out can be backfilled with waste rock, or tailings, often with the addition of cement to strengthen the backfill pumped as a slurry from the surface. Temporary supports may also be used to allow the pillars to removed ("pillar robbing"): once work in an area is completed the temporary supports are removed and the roof is allowed to collapse.



For "weak" ores, other forms of support (in addition to, or as alternatives to the pillars) may be required to ensure a safe working environment. This supplementary support is then removed as the working face proceeds, allowing the roof to collapse in the worked out area. This is known as longwall mining.



Ore deposits that dip steeply or which extend over a great vertical distance are mined in such a way as to make use of gravity to transport ore to one or more central haulage points within the mine and in some cases to break the ore under its own weight. A mining method called stoping is used to extract the ore in sequential blocks. The cavity formed is called a stope, with subsequent ore being blasted from the roof to the stope floor and transported to an ore haulage area via ore passes (vertical or nearvertical channels).

In-situ

Also sometimes known as solution mining, this method involves the synthesis of expertise from a number of disciplines including hydrometallurgy, hydrology, geology, geochemistry, rock mechanics, chemistry and environmental management (Hiskey, 1994). In-situ mining is accomplished by the injection of a suitable leaching agent into a porous and permeable mineral deposit. During its passage, the leaching agent (normally water-based) dissolves the metal(s) of interest. The metal-laden leachate is then pumped to the surface where it is purified and further processed to recover a saleable product. At present its use is relatively limited, particularly outside the recovery of uranium from permeable sandstone deposits. It has, however, been used to recover copper from oxide mineral remnants in underground and open pit workings and from caved-in stopes. In-situ leaching has certain advantages over conventional mining and milling, including lower capital investment, lower operating costs, and faster start-up times.

Potential releases, effect and impacts

There are two major releases from surface and underground mining, first those associated with mineral wastes (i.e. overburden and waste rock) and second the generation of contaminated waters. Other releases may be locally significant depending on the characteristics of the site and external environment, for example noise, vibration and dust (from blasting and road haulage), surface subsidence, diesel-related fumes (from mechanical plant), water contamination by nitrates (arising

from ammonium nitrate used in blasting) and visual disturbance from operational activities, waste disposal and ore extraction.



Mine water is generated when water collects in mine workings as a result of inflow from rain or surface water and from groundwater seepage. During the active life of the mine, water is pumped out to keep the mine relatively dry and to allow access to the ore body for extraction while surface water is controlled using engineering techniques to prevent water from flowing into the mine. Pumped water may be used in extraction and beneficiation activities (including dust control), pumped to tailings impoundments, or discharged as a waste. The quantity of mine water generated varies from site to site, and its chemistry is dependent on the geochemistry of the ore body and the surrounding area. Water exposed to sulphur-bearing minerals in an oxidizing environment, such as an open pit or underground workings, may become acidified and contaminated with metals (e.g. acid rock drainage). Acid drainage may also be generated by sulphide-rich waste rock piles that are permeable to both air and water.



Acid rock drainage is widely considered to be the most serious environmental problem caused by the mining of sulphide ore deposits (see Summitville case study, below). It results from the reaction of iron sulphides with oxygen and water. The sulphides oxidise to generate acidity and other chemical species that are capable of oxidising other metal sulphides into water-soluble sulphate form. The resulting water can be both acidic and highly laden with iron and other metals. The oxidation of the iron sulphides is catalysed by naturally occurring bacteria which thrive in the acidic conditions. When mining ceases and the pumps are removed, water levels can rebound to natural levels, often producing a large volume of contaminated water, which is continually replenished as fresh (oxygenated) water enters the workings (and contaminated water leaves).



If acidic drainage is left untreated it can contaminate groundwater and local watercourses, restricting water use and damaging ecosystem and human health. The drainage can be treated, but this is expensive and preventative approaches are more cost-effective.



The main release from in-situ mining is contaminated water which may contain a wide range of dissolved species from the ore (similar to acidic drainage) and chemical species added to the leaching solution. Although there are other technical limitations to the use of in-situ mining, the major unresolved issue is the question of losses of metal-laden leachate and subsequent contamination of groundwater resources. Experience with heap and dump leaching (surface-based variants of in-situ mining, see below) indicates that there are good reasons to be cautious about extending the use of in-situ mining until this issue is fully resolved.

Innovations in mining

The major thrust of innovation in mining has been the development of open pit techniques and exploitation of the economies of scale created by developing larger mining equipment and largerscale extraction techniques. For example, major developments have included (ILO, 1990): improved ventilation equipment and improved visibility; larger and more efficient hoisting machinery, compressors, pumps, etc.; improved mine development equipment and procedures for mechanised shaft sinking, shaft freezing, raising and drifting; larger and heavier equipment with better drill bit and blasthole drilling techniques; trackless mining equipment, initially to mechanise ore loading and transport; improved blasting agents and techniques for greater efficiency and safety; electric mining shovels and draglines with significantly greater bucket capacities; very large (200 ton) diesel electric haul trucks and front loading equipment; portable and mobile in-pit ore crushing and overland conveyor systems, allowing more continuous mining operations and the elimination of truck haulage where hauls are long and adverse; self-propelled scalpers and scrapers to integrate extraction and transport operations; on-line/automated production monitoring and control to optimise extraction and truck and shovel capacity utilisation through scheduling/dispatching and route selection; and computerised and remote control systems for underground train haulage and mine pumps.



In underground mining, advances have also been made in the use of "right-in-space" mining methods (Almgren et al, 1996), which seek to ensure that drill holes, stopes and other underground workings are placed more accurately in relation to the orebody during both development and production. This allows high ore recovery and reduces dilution by unwanted gangue, reducing waste that requires disposal and management.

Mineral Processing

Introduction

Mineral processing is defined as the physical processing of minerals. It does not result in any chemical changes to the mineral components of the ore, but is a means of achieving the physical separation (and concentration) of different mineral phases (e.g. target minerals from gangue minerals or of one valuable mineral from another) (Hayes, 1993). For non-metal mineral resources, mineral processing can produce a final product, but in the case of metals it is an intermediate stage as it does not affect the chemical combination of the metal with oxygen, sulphur and so on. Mineral processing is normally an intermediate stage between mining and extractive metallurgy, although there are exceptions: for example heap and dump leaching of "as-mined" ore. The outputs from mineral processing (concentrates) form the inputs to extractive metallurgy (either hydrometallurgical or pyrometallurgical processes, see Extractive Metallurgy, below).



Mineral processing methods can be divided into two groups: size reduction and separation of mineral phases. Both these stages involve varying degrees of screening and classification. Dewatering is also used at various stages to prepare intermediate products for subsequent stages, or to control the water content of the final products.

Size reduction

Size reduction is undertaken using crushers (e.g. jaw, gyratory and cone crushers) and grinding mills (rod, ball, hammer and impact mills). Different types are employed depending on the planned throughput (tonnes per hour), ore characteristics (e.g. friable, "sticky", "elastic") and the degree of size reduction required. Crushing is used to reduce incoming ore from boulder-size down to 25mm or less. This material may then pass to grinding mills or may be processed directly if the valuable mineral is sufficiently liberated. Grinding reduces particle size, down to a lower limit of about 10μm.5 In contrast to crushing, which is carried out on run-of-mine ore, grinding is almost always done wet and the output from grinding is a slurry (a suspension of fine particles in water).



Size reduction can be realised via three routes:

• Slow compression, leading to a range of coarse particles.


• Fast compression or impact, leading to a fine-coarse particle size range.


• Abrasion, leading to very fine particles and coarse particles that are being abraded.

Crushing and grinding (in tandem with sizing) are used to liberate economic and non-economic minerals from one another and thereby produce suitable feeds for subsequent processes in which separate mineral phases can be generated. However, grinding does not normally produce a completely liberated mineral product and some particles may be mixture of two or more mineral species, with physical or chemical characteristics representative of that mixture. This can result in target minerals being disposed of with the gangue minerals, or the dilution of the mineral concentrate by gangue minerals. Separation and regrinding of "mixed" particles to a finer size is often used to improve liberation, while avoiding overgrinding of particles that are already liberated. However, grinding (and crushing) are energy intensive and there is a trade-off between the cost of additional grinding and the value of the extra mineral likely to be recovered. Grinding accounts for nearly half of the total energy used in producing a mineral concentrate by mineral processing (Rankin and Wright, 1992). Energy consumption in mineral processing can, therefore, be reduced by optimising the process flowsheet to minimise regrinding. Of particular importance to optimisation is the capacity to run the plant at steadystate irrespective of changes in feed mineralogy and grade (Rankin and Wright, 1992). Recycled streams (also known as circulating loads) at all stages of the mineral processing flowsheet plant are, however, used to smooth out changes in feed characteristics that occur periodically, and therefore the need to minimise recirculating material through the grinding mills must also be balanced against the stability of the overall circuit.



Characteristics of the major crusher and grinding mills are shown in Table 4.

Table 4 Characteristics of major crushers and grinding mills (after Wills, 1997)

Screening

Screening is used to separate relatively coarse material (+250μm) and is typically implemented to reduce excessive crushing and overgrinding, to prepare a narrow size range for subsequent processes, or a closely sized end product. Generally, it produces little waste as it operates in a closed circuit with up and downstream processes. However, noise and airborne dust may be generated, and energy is expended.

Classification

Classification involves the separation of mixtures of minerals into two or more products on the basis of the velocity with which mineral grains fall through a fluid medium. Classification is used for fine particles that cannot be efficiently screened. Typical unit processes include hydraulic classifiers, mechanical classifiers (rake and spiral) and hydrocyclones (see Size separation, below).

Separation of mineral phases

Mineral separation can be achieved by employing differences in mineral characteristics, based on:

• Particle size.


• Particle density.


• Magnetic properties.


• Electrical properties.


• Surface chemistry characteristics (flotation).

However, no separation is perfect due to the following factors:

• Mixed gangue/target mineral particles (middlings).


• Physical entrainment.


• Overlapping physical or chemical characteristics of mineral phases.

The separation technique of choice is based upon a number of site-specific factors, including the size of the operation, particle size at which minerals are sufficiently liberated, mineralogy, the overall processing circuit (from mining to extractive metallurgy) and so on. Ultimately, each process produces a concentrate (rich in the target mineral or minerals), tailings (containing mainly sub-economic or non-economic minerals) and middlings (particles in which target and gangue minerals have not been

fully liberated). Middlings are normally reground or sent to an alternative process to liberate or otherwise recover the valuable mineral, while tailings are disposed of as waste. Of the various separation processes, flotation is now the dominant method for the production of mineral concentrates, particularly from sulphide ores.

Particle size

Different minerals have a variable resistance to crushing, abrasion and deformation during crushing and grinding, which can therefore result in "soft" minerals being reduced to smaller sizes than "harder" minerals. Consequently, by the use of sizing it is theoretically possible to begin the process of separation and concentration of the valuable mineral(s). However, sizing is more normally used to split streams with a broad size range into two or more products with a narrower size ranges. These more narrowly sized products are then fed to subsequent processes.



Standard separation by size involves sieves, screens, spiral/rake classifiers and thickeners. Classifiers and thickeners work on the principle that in water a larger particle will drop faster, therefore the sediment will contain coarse particles, while the water above it will contain smaller particles (still in suspension). Subsequent separation of the sediment and water (as a continuous process) therefore allows particles to be separated by size. However, differences in particle density (e.g. base metal minerals are typically denser than the major associated gangue minerals) results in small dense particles having similar settling characteristics to large light particles. Therefore a certain amount of material is misplaced during the separation process, and this can be a source of inefficiency in subsequent processes and increased environmental releases.



Separation by size is increasingly conducted using hydrocyclones, which apply centrifugal force to accelarate the settling process. Hydrocyclones require less area than other size separation techniques and by placing them in parallel, very high volumes of slurry can be sized. However, as for the other size separation techniques, different particle densities result in misplaced material (i.e. the coarse fraction may contain small, but dense, particles).

Particle density (gravity concentration)

The accelerated settling of dense particles relative to lighter particles is used in the process of separation by density. Typical methods employed include dense medium separation, jigs, shaking tables and spiral concentrators. Jigs and shaking tables employ mechanical vibration to accelerate the separation process. As is the case for separation based on size, the similar behaviour of light large particles and small dense particles results in misplaced mineral. However, this can be minimised by using density separation on narrowly sized feeds.



In gravity concentration, coarse separation enables a higher tonnage throughput, therefore regrinding/classification is used to remove mineral in as coarse a state as possible.



Gravity concentration has declined in importance due to the advent of froth flotation (see below). When gravity is applied over flotation, it is normally on the basis of cost, as the minerals liberated at a size above those in the normal flotation range can be concentrated more economically using gravity methods. On the plus side, gravity methods do not have the cost of reagents (as does flotation), are relatively simple to control and produce comparatively little environmental pollution. However, a lack of uniformity in feeding results in substantial failure in operating efficiency and may lead to losses in recovery.



Heavy or Dense Media Separation relies on the principle that a thick suspension or pulp of some heavy material (e.g. ferrosilicon) in water behaves as a heavy liquid and is normally applied as a preconcentration stage to reject coarser gangue prior to grinding for final liberation. It is also applied in coal preparation to produce a graded end product.



Gravity concentration works most efficiently when treating feed with relatively narrow size ranges, e.g.:

• Sand tables (100μm-3mm).


• Slimes tables (+10μm).


• Jigs (3-10mm).


• Pinched sluices and cones (mineral sands).


• Reichert cones (100-600μm).


• Spirals (75μm- 3mm).

Pneumatic tables (air) may also be used for cleaning heavy mineral sand concentrates to remove silica or for the purification of light industrial minerals (e.g. vermiculite).

Magnetic properties

Selection of different mineral phases can also be achieved by employing the differing magnetic properties of minerals. Broadly speaking, minerals fall into two groups: diamagnetic (those minerals repelled by magnetic fields) and paramagnetic (those minerals attracted by magnetic fields). Separation can be undertaken using high and low intensity magnetic fields, but is limited to ores in which there is a demonstrable difference in the magnetic properties of the target and gangue minerals (i.e. it is not applicable where all the minerals are diamagnetic).

Electrical properties

Utilising the forces acting on charged or polarised particles, conducting and non-conducting minerals can be separated. As in the case of magnetic separation, this approach is limited to a relatively small number of mineral species.

Flotation

those that physically stabilise the froth (to ensure that the bubbles do not burst before the froth isFlotation involves the passing of air bubbles through a dilute slurry of fine mineral particles held in large cells. Minerals that are hydrophobic ("water-hating") become preferentially attached to the air bubble and float to the surface of the slurry to form a froth, where there can be removed. Hydrophilic ("waterloving") minerals remain in the bulk slurry. The key activity in flotation is the control, alteration or accentuation of the hydrophobic or hydrophilic characteristics of target and gangue minerals. In broad terms, two options are possible with flotation: separation of valuable minerals from gangue minerals and the separation of one valuable mineral from another. During the flotation process minerals may be altered from hydrophobic to hydrophilic by the use of chemicals termed depressants, which alter the surface chemistry of the mineral in question. Hydrophilic minerals may be made hydrophobic by the addition of other reagents known as activators. Other chemicals added during the processing include removed), those that adjust the pH of the slurry (to aid in the control of selectivity for one mineral over another) and those that ensure that mineral particles do not aggregate (dispersants).



Flotation technologies have been used since the 1920s, but it was not until the early 1980s that automated control systems became available which enabled a fine-tuning of the flotation process to optimise its performance. Even so, with a complex polymetallic sulphide ore treated by flotation it is not unusual for 20-30% of the contained value to be lost during mineral separation (Wills, 1997).



Flotation reagent are now more expensive in relative terms, and consumption has increased as surface areas of the mineral phases has increased (due to more complex ores and finer grinding sizes).

Dewatering

Dewatering is accomplished by either coagulation (the "sticking together" of extremely fine particles), flocculation or through the natural settling characteristics of particles (gravity sedimentation, centrifugal sedimentation) and filtration. Inefficiencies include the loss of fine particles and losses of the chemicals added for the purpose of promoting flocculation or coagulation. Coagulation and flocculation can be driven by "simple" reagents such as lime or sulphuric acid in some cases, but increasingly they are undertaken with natural organic polymers (e.g. starch, gelatine) and more likely synthetic materials, loosely termed polyelectrolytes (e.g. polyacrylamide).



Dewatering is used for drying of final concentrates or intermediate stages to prepare feeds for the subsequent processes. The variation in water content at various stages of processing is shown on Table 5.



Table 5 Water content during mining and processing of ores (after Wills, 1997)

Potential releases, effect and impacts

Many of the unit operations within mineral processing are of limited interest in terms of their potential to impact the external environment as they operate within what are, effectively, closed circuits (i.e. many unit operations pass effectively all of their input to the next process, and therefore they involve only very limited transfers of materials from the process to the external environment, such as dust, noise and vibrational energy).6 However, this is not true for flotation, which is the biggest single source of tailings in mineral processing.



Tailings generated during flotation are mainly composed of finely sized gangue minerals within which are contained varying amounts the target mineral(s). The concentration of target minerals in the tailings is mainly related to process economics, process efficiency and mineralogical constraints. The tailings are commonly discharged in the form of a slurry to tailings impoundments, which may cover tens of hectares and present a serious problem in terms of dusts in drier climates. If the tailings are rich in sulphides (particularly pyrite), impoundments can also generate significant volumes of acidified water, which may require collection and treatment during the operation of the impoundment and after its closure. Other associated effects include those arising from the chronic discharge of suspended solids, the leaching of associated flotation chemicals and failure of impoundment walls.



Tailings, by virtue of their semi-liquid composition, are difficult to secure, particularly over long periods of time extending beyond the active life of the mine. Two aspects of tailings disposal are crucial:

• Dam stability, which requires close control of the separation of coarse and fine tailings, installation of adequate drainage, control and monitoring of seepage, geotechnical control of the dam slope, and filtering mechanisms; and


• Risk analysis, leading to the building of extra capacity and strength into the reservoir to withstand freak earthquakes and weather events. It was the failure to undertake the proper risk analysis - a direct reflection of cost-cutting and lack of regulation - which accounted for many disasters associated with tailings in, for example, Papua New Guinea and Peru. Indeed, it was the relatively liquid composition of the tailings from the Ok Tedi and Porges Project in Papua New Guinea, combined with the lack of storage sites, which led to the tailings from that mine’s being disposed of (after treatment) in the local river, resulting in serious environmental degradation.

Table 6 shows the "idealised" costs associated with processing and related activity. This table is based on a 500 td-1 concentrator, although these costs should not be taken as "typical" as these would be impossible to properly establish given that site characteristics vary to such a large extent.

Table 6 Costs associated with processing and related activity (after Wills, 1997)

Extractive Metallurgy

Extractive metallurgy can be subdivided into two major disciplines, namely hydrometallurgy and pyrometallurgy. A third discipline, electrometallurgy is not considered here in detail as its use is relatively limited in the mining sector (e.g. mainly the production of aluminium and some zinc).



Hydrometallurgy involves the use of solvents (normally water-based) to dissolve the metal(s) of interest, producing a dilute, metal laden solution which is then further processed to recover the metal. Typically, hydrometallurgical processing involves temperatures below 50°C (although some pressurised leaching systems do operate up to 250°C (Rankin and Wright, 1992)). In contrast, pyrometallurgical processes are normally operated at temperatures in excess of 800°C.

Hydrometallurgy

Hydrometallurgical methods of ore treatment are most commonly used for gold, uranium, aluminium and copper and to a lesser extent zinc and nickel. In particular, ores containing oxide material (about 10 per cent of non-ferrous ores) are amenable by hydrometallurgical routes such as leaching. Ore is first crushed and processed according to the requirements of the subsequent processes (e.g. limited crushing for heap and dump leaching, processing to concentrate-stage for vat leaching). A leaching agent (lixiviant) is then used to extract the valuable metal(s) in the form of a dilute, metal-laden solution. This solution then passes to the metal recovery stage, which may involve precipitation, solvent extraction or electrowinning. In a strict sense, electrowinning should be classed as an electrometallurgical process, but as it is normally associated with leaching, it tends to be grouped with the other hydrometallurgical processes.

Leaching reagents and solvents

Acids and oxidants

Acid leaching of ores and concentrates is the most common method of hydrometallurgical extraction, particularly for the recovery of copper. Typical acidic leaching agents include hydrochloric acid (HCl), sulphuric acid (H2SO4) and ferric sulphate (Fe2(SO4)3). Oxidised copper minerals such as azurite, malachite, tenorite and chrysocolla are completely soluble in sulphuric acid at room temperature. Other, less oxidised minerals such as chalcocite, bornite, covelite and chalcopyrite require the addition of ferric sulphate and oxygen (as oxidants) to accomplish leaching.

Alkalis and ammonia-based reagents

For certain copper minerals, alkaline (or basic) leaching is more effective. Alkaline leaching is more selective than acid leaching and particularly appropriate for ores with large amounts of acidconsuming carbonate rocks. However, this selectivity often results in lower recovery if the metals are not fully liberated during crushing and grinding. Silica- and silicate-rich ores can be treated using alkaline leaching agents at raised temperatures. The principal reagents used in alkaline leaching are the hydroxides and carbonates of sodium and ammonia, but potassium hydroxide and calcium hydroxide are also used. Those metals which form amines (e.g. copper, cobalt and nickel) can be dissolved in ammoniacal ammonium carbonate or ammoniacal ammonium sulphate solutions at atmospheric pressure.

Bacterially-mediated leaching

This is applied to low-grade sulphide ores in dump and heap leaching operations and has revolutionised parts of the mining industry in recent years by enabling economic processing of what were previously considered wastes. Leaching is much slower than typical acid or basic leaching and relies upon the capacity of bacteria such as Thiobacillus ferrooxidans and Thiobacillus thiooxidans to oxidise ferrous iron to ferric iron (which in turn oxidises other metal sulphides, producing water-soluble sulphates). Sulphuric acid is also a product of the bacterial activity. The main requirements for bacterial activity are oxygen, ammonia, nitrogen, phosphate, a suitable temperature (approximately 30°C) and acidity (approximate pH of 2.0). Higher or lower temperatures (5°C or 50°C) or pH (0.5 or 4.5) do not tend to kill the organisms, but instead dramatically reduce their activity.



Bacterially-mediated leaching is also used to process refractory gold, hitherto unrecoverable due to its crystalline association with pyrite (which the bacteria can readily dissolve). Advances in biotechnology combined with the environmental and economic advantages which bacterial leaching technologies appear to have over other larger scale, more capital intensive and more polluting traditional processes - like reverberatory furnace smelting - may herald substantial changes in the structure of the minerals industry (Warhurst, 1992), although recent improvements in SO₂ capture technology in smelters are undermining the potential competitive advantages of hydrometallurgy.



Although to date bioleaching has only been applied commercially to the recovery of gold, uranium, copper, and nickel, it has also been suggested as a route to effective heap leaching of low grade zinc ores. However, to follow this route further advances are required in the solvent extraction of zinc from iron-contaminated liquors. Electrowinning of zinc, while more complex than is the case for copper is technically feasible (Grant, 1994). The capacity to leach low-grade zinc ores would improve overall resource recovery.



Cyanidation

Cyanide (as a sodium or potassium cyanide solution) is used to dissolve gold. Alternatives to cyanide do exist, but few have been used commercially (see Cyanide Case Study, below). One succesful example is Newmont Gold’s innovative approach to bioleaching, which combines bio-oxidation with a patented ammonium thiosulphate treatment as an alternative to cyanidation for refractory ores (Warhurst and Bridge, 1996).



Amalgamation

Mercury is used in the amalgamation of gold. This practice is widely applied at small-scale mining operations, principally in developing countries and causes considerable damage to ecosystem and human health (see below).

Leaching methods

Dump leaching

Dump leaching refers to leaching that takes place on an unlined surface. The term "dump leaching" derives from the practice of leaching materials that were initially deposited as waste rock; however, now it also is applied to of run-of-mine, low-grade sulphide or mixed grade sulphide and oxide rock placed on unprepared ground specifically for leaching. Copper dump leaches are typically massive, with waste rock piled into large masses ranging in size from 20 feet to over 100 feet in height. These may cover hundreds of acres and contain millions of tons of waste rock and low-grade ore (Biswas and Davenport, 1976).



Heap leaching

In contrast to dump leaching, heap leaching refers to the leaching of low-grade ore that has been deposited on a specially prepared, lined pad constructed using synthetic material, asphalt, or compacted clay. In heap leaching, the ore is frequently pre-treated using size reduction (e.g. crushing) prior to placement on the pad. Site-specific characteristics determine the nature and extent of the crushing and the leaching operations used (USEPA, 1989). Cyanide heap leaching is a relatively inexpensive method of treating low-grade gold ores while vat leaching is used for higher grade ore. Heap leaching is generally used to treat ores containing less than 0.04 ounces of gold per tonne. Tank or vat methods, are generally used to treat ores containing more than 0.04 ounces of gold per tonne. The cut-off value for the different treatment routes is dependent on many factors, including the price of gold and an operation's ability to recover the precious metal (van Zyl et al, 1988).



Vat leaching

For copper, the vat leaching process works on the same principles as the dump and heap leaching operations except that it is a high-production-rate method conducted in a system of vats or tanks using concentrated lixiviant solutions. It is typically used to extract copper from oxide ores by exposing the crushed ore to concentrated sulphuric acid. The vats are usually run sequentially to maximise the contact time between the ore and the lixiviant (USEPA, 1989). Vat leach units may be large drums, barrels, tanks or vats. The design capacity of the leaching units is dependent on the amount of ore to be leached. For example, a 25-meter-long, 15-meter-wide, and 6-meter-deep vat unit is capable of leaching between 3,000 and 5,000 tons of ore per cycle. Vat leaching of concentrates (as opposed to ore) is also applied (e.g. the cyanidation of gold-rich sulphide concentrates).

Metal recovery

For copper there are two main methods of recovering the metal from dilute aqueous solutions: cementation onto scrap iron and solvent extraction/electrowinning.



Cementation (precipitation)

Typically, cementation precipitators are shallow-round or stair-stepped wooden or concrete basins (U.S. Congress, Office of Technology Assessment, 1988). The simplest and most common precipitation system used in the copper mining industry are open-launder-type cementation systems where copper-rich solution flows through a trough filled with scrap iron. Metallic copper precipitates on the iron (with a concomitant dissolution of the iron). The copper precipitate is invariably contaminated

with iron and requires further refining before it can be sold.



Solvent extraction and electrowinning (SX/EW)

For copper, the solvent extraction operation is a conducted in two stages. In the first, dilute and impure leach solution containing dissolved copper, iron and other base-metals (from the leaching stage) is passed to a mixer for extraction of the copper. In the mixer, the aqueous solution is contacted with an active organic extractant (chelating agent) in an organic diluent (usually kerosene), forming a copperorganic complex. The extractant is designed to selectively extract only the desired metal (in this case copper), while impurities such as iron are left behind in the aqueous phase. The organic phase (containing the organic-copper complex) is then physically separated from the barren aqueous phase (Anonymous, 1991). The latter is recirculated back to the leaching units while the copper-loaded organic phase is transferred to the stripping section where the copper is removed by mixing the organic

phase with concentrated sulphuric acid solution (spent electrolyte from the electrowinning stage) to produce a clean, high-grade solution of copper for electrowinning. Copper is then plated out of solution onto inert (non-dissolving) cathodes made of lead alloyed with tin and calcium or of stainless steel. The obvious advantage of solvent extraction is that cathode copper of saleable quality can be produced directly from leach solutions and therefore further purification is not required. However, there are signs that traditional SX/EW for copper may be under threat from systems that remove the need for SX by allowing direct recovery from low-concentration copper solutions (Clifford, 1997).

Potential releases, effects and impacts

Releases from heap and dump leach piles during and after closure

When heap leach operations are concluded, a variety of constituents remain in the wastes. These include cyanide not removed during rinsing or neutralization (for gold leach operations), acid or alkalis (for base metal operations), as well as heavy metals, sulphides and other metal-bearing minerals. After the operation has been closed or reclaimed, in the absence of proper design and control measures, runoff from the spent ore may occur. This runoff may contain constituents associated with the ore, such as heavy metals, and suspended solids. Depending on the method and completeness of detoxification, leachate from spent ore may also maintain a high pH (gold leaching) or low pH (copper leaching) over an extended time period.



Releases from active heap and dump leach units

The release of leaching agents (e.g. cyanide in gold operations, sulphuric acid in copper operations) from active leach piles or leachate collection ponds may occur during snowmelt, heavy storms, or failures in the pile or pond liners and associated solution transfer equipment,9 with severe implications for surface and groundwater quality.



SX/EW sludge

Sludge is the semi-solid gelatinous materials that can accumulate in solvent

extraction/electrowinning tanks. These sludges are colloids of suspended material that cannot be easily settled or filtered. The solvent extraction process generates a sludge consisting of a solid stabilized emulsion of organic and aqueous solutions. It is located at the organic/aqueous interface in the settlers and is periodically removed from the system, and centrifuged or otherwise treated to remove the organics. The aqueous solutions and the solids are disposed of and the organics are returned to the solvent extraction circuit for reuse. Depending on the characteristics of the ore body, the sludges may contain base or precious metals in quantities sufficient for recovery.



Spent Electrolyte

Spent electrolyte is generated during electrowinning activities. Historically, electrolyte went through a stripping step and was subsequently discharged to a tailings pond. Today, due to economics, this effluent is recycled to reduce capital costs associated with the electrolytic acids used in these operations. Over time, electrolyte in the electrowinning cells becomes laden with soluble impurities and copper. When this occurs, the solution is removed and replaced with pure electrolyte (to maintain the efficiency of the solution and prevent co-precipitation of the impurities at the cathode). Purification of the spent electrolyte is done by electrowinning in “liberator cells”. Liberator cells are similar to normal electrolytic cells, but they have lead anodes in place of copper anodes. The electrolyte is cascaded through the liberator cells, and an electric current is applied. Copper in the solution is deposited on copper starting sheets. As the copper in the solution is depleted, the quality of the copper deposit is degraded. Liberator cathodes containing impurities (such as antimony) are returned to the smelter to be melted and cast into anodes. Purified electrolyte is recycled to the electrolytic cells. Any bleed electrolyte is usually neutralized with mill tailings and disposed of in a tailings pond (USEPA, 1984).



Mercury releases

A growing area of concern relates to mercury toxicity resulting from its amalgamation with gold metal as a recovery method in small-scale gold mining in developing countries. The small-scale gold miner often works for himself, putting in long hours under hazardous conditions (e.g. often standing in insect/disease-ridden waters). The miner frequently moves from plot to plot in nomadic fashion without a long-term perspective on the environmental impacts of his activities. These conditions contribute to the excessive use of mercury reagents to amalgamate the mercury in the erroneous belief that the greater the amount of mercury used the greater the rate of gold extraction (CETEM, 1990). However, it is a precise ratio between mercury and gold that should be sought to maximise gold recovery. A number of countries have sizeable small-scale alluvial gold mining sectors, including Brazil, Colombia, Ecuador, Ghana, Papua New Guinea, Peru, Philippines, and Zimbabwe. In each case, the volume of gold produced by such means is estimated to exceed 15 tonnes per year, of which a combined total of 100 tonnes are produced using mercury. With an estimated usage rate of 4 kg of mercury per kg of gold produced, the total environmental loading of mercury from such sources is roughly 400-500 tonnes per year. It is the release of this excess mercury into both water systems and the atmosphere which accounts for the fast spreading incidents of mercury poisoning which are being reported in many mining regions of gold-producing countries. In tropical climates, mercury also tends to have a higher methylisation rate (involving the reaction of elemental mercury to form easily absorbed and highly toxic organic compounds). Furthermore, the detection and control of the problem is made more difficult by the complexities of the food chain and the fact that affected fish can be eaten hundreds of miles downstream from actual mining activities.



Unlike cyanide, mercury does not degrade in the environment, but is transformed into organomercury species that are both toxic and bioaccumulators (i.e. they tend to accumulate in the food chain). There are a number of safer, cleaner alternatives to mercury use, for example gravity separation and cyanidation and simple procedures to reduce losses of mercury to the environment during gold recovery (e.g. closed retorting). However, there are many reasons for the dominance of mercury in the small-scale sector: mercury is relatively cheap, it is simple to use and requires limited

capital outlay on equipment, there is often a lack of education on the part of the users as to the health and environmental implications, and the sector is largely unregulated, or where regulation has been enacted, enforcement is not possible due to lack of resources, civil unrest or other socio-political factors. The question of how to reduce the use of mercury in the small-scale mining sector is one that requires input from social scientists, politicians, economists and engineers if viable and lasting solutions are to be found.

Pyrometallurgy

Pyrometallurgical processes are currently the backbone of the recovery of copper, zinc, nickel and lead from sulphide deposits. The process route includes mineral processing (normally flotation) to produce a concentrate, followed by smelting, which breaks down the crystalline structure of the minerals by heat-fuelled oxidation. This produces a matte (containing up to 40 per cent metal) which in its molten form is converted and separated into blister (about 97-99% pure) and an iron-silicate slag - which may have sufficient economic value to be worth processing. This waste is deposited in "tailings ponds" which, as noted above, are a potential source of toxic leakage and dust pollution if not carefully contained and managed. Since blister metal is too impure for most industrial applications, refining is normally necessary. This is usually undertaken by a fire process (using a reverberatory furnace) if the feed has a low by-product content, or by electrolysis if additional metals are to be recovered. The resulting cathodes which are usually approximately 99.8-99.9% pure are marketed directly to the semifabricators, or cast into shapes (e.g. wire bar).



Potential releases, impacts and damage

Smelting gives rise to four potentially pollutant products: waste gas, fugitive gas, smelter dust and effluents. The extent of damage can generally be related to four variables: first, the overall efficiency and type of the smelter process (smelters may last for decades and many particularly polluting ones date back to the 1940s and 1950s); second, the environmental impact controls imposed upon the smelter and implemented by its operators; third, the levels of naturally occurring impurities in the copper or metal concentrates (e.g. arsenic and lead, typical of Peru; bismuth, typical of Bolivia, and molybdenum, typical of Chile); and fourth, the effect of smelter location and geographical features on emission controls. Two serious problems are sulphur dioxide emissions, and losses of toxic volatile elements such as arsenic (see Arsenic Case Study, below).



Sulphur dioxide emissions and acid rain

World-wide, the smelting of copper and other non-ferrous metals releases an estimated 6 million tons of sulphur dioxide into the atmosphere each year - which constitutes 8 per cent of the total emissions of the sulphur compounds that cause acid rain (Young, 1992). Such emissions produce "dead zones" where little or no vegetation survives. Such an area around the Ontario nickel smelters (of INCO and Noranda) measures 10 400 hectares and acid fallout from the smelters has destroyed fish populations in lakes 63 kilometres away. In the United States, the dead zone surrounding the Copper Hills smelter in Tennessee covers 7 000 hectares.



The first difficulty in determining emission factors from smelting relates to the complexity of the relationship between sulphur dioxide emissions and local sulphate concentrations, the latter being largely responsible for the associated environmental hazard.11 For example, when sulphur dioxide is emitted into the atmosphere it reacts with water to form a sulphuric acid aerosol and with various metallic ions to form metallic sulphates, which are of respirable size range. The rate of oxidation of sulphur dioxide to various sulphate related species varies between 0.1 and 30 per cent per hour (i.e. by a factor of 300). Moreover, the conversion is catalysed by photochemical smog and varies with sunlight intensity, ambient temperatures, humidity and the presence of particulate matter (Sawyer, 1977). The suggestion that health problems are the result of sulphates rather than sulphur dioxides has dramatic implications for sulphur dioxide discharge policy, since there is little relationship between sulphur dioxide discharge and local sulphate concentrations, which have more to do with other ambient pollutants, prevailing winds and humidity levels; the related subject of acid rain therefore becomes important.



The most comprehensive documentation for pollution from smelting contributing to acid rain is for North America, where in 1980 SO₂ emissions were 28.9 million tons of which 3.5 million tons (or 12 per cent) - originated from non-ferrous smelters. The nickel and copper smelting facilities in Canada of INCO at that time were the continent’s largest single source of air pollution, and facilities such as the Falconbridge and Kidd Creek smelters followed close behind accounting alone for more than 19 per cent of North American acid rain. Ontario in Canada is unique in that it is a significant source and recipient of acid rain, due in the main to the processing of local nickel, copper and iron ores. In addition, the rock formation which provides this mineral wealth is the granitic Pre-Cambrian Shield, which is characterised by thin acid soils which provide little chemical buffering to counteract the incoming acid from the atmosphere. Consequently acid rain damage to the natural environment in Ontario is severe and widespread. Similar damage is found in the granitic Andes regions which host the mineral wealth and smelting facilities of Chile, Peru and Bolivia. The problems in the latter region are however less well documented.



Innovations in Pyrometallurgy

Widespread concerns over the effects of smelter emissions during the last two decades, particularly with increasing availability of research data on the acid rain phenomenon, combined with economic concerns over energy availability and cost, and low efficiency levels, have brought smelter design to the forefront of recent innovation efforts in the mining industry. Technical change resulting in a cleaner continuous smelting process, often with electric furnaces replacing reverberatory furnaces, is at the heart of much innovation in lead, copper and tin production. Closed systems avoid SO₂ emissions while high SO₂ concentrations in the effluent gases enable its recovery in the form of sulphuric acid. At the same time, such systems are often more efficient in their use of energy and enable higher recovery of metals and by-products.



The pyrometallurgical industry also has within its power the opportunity to, using existing technologies, design and produce wastes with specific characteristics (Coppin et al, 1995). This could entail minimising the concentration of potentially hazardous components, or adjusting the physical and chemical characteristics of the wastes to endow them with greater resistance to detrimental changes once disposed of to the external environment, thereby reducing disposal costs. The wastes could also

be designed as a resource (e.g. use as an aggregate or for potential recovery of contained values at some future time).



Considerable progress towards pollution prevention in the smelting industry has been made over the last few years through the redesign of the production process for sulphide ores to facilitate sulphur dioxide capture and its efficient conversion to sulphuric acid. Together with a steep rise in energy prices during the 1970s, the demonstration of the linkage between sulphur dioxide emissions and acid precipitation challenged the smelting industry to find ways of reducing sulphur dioxide emissions while continuing to be viable in a very competitive world market. In seeking to meet the challenge posed by competitive market pressures and regulatory and societal demands for better environmental performance, new technologies have improved process efficiency and cut emissions by reducing the number of stages in the smelting process, increasing the concentration of sulphur in the off-gas, and enclosing the process so as to make the capture of off-gases as efficient as possible. Noranda Minerals Inc, for example, has reduced SO2 emissions at its seven metallurgical facilities from 800,000 tonnes per year in 1970 to 270,000 tonnes per year in 1990 by adopting smelter technologies that reduce SO2 production, and by increasing the conversion to sulphuric acid which is sold as a byproduct (Noranda, 1990).

Copper

For copper, the traditional three-phase roasting, converting and smelting process is currently being replaced by a combined step-direct matte smelting and continuous smelting using fluid bed roasters instead of multiple hearth roasters. The concentrate is suspended in a stream of hot air and flue dusts are recovered for re-processing. Closed systems avoid SO₂ emissions and electric or electronic furnaces instead of reverberatory furnaces are also used. The advantages over the latter include: high SO2 concentrations in the effluent gas enabling efficient recovery as sulphuric acid; greater energy efficiency and reduced fuel consumption; reduced blowing time in the converter and higher throughputs. The most efficient "flash" smelter currently in use is the Outokumpu process where the concentrate is dispersed in an oxygen stream. Of the continuous copper-making processes, the Noranda and Mitsubishi processes are considered state-of-the-art, although recently the Isasmelt process (Australia) and Cyclomelt process (Holland) which came into production in the late 1980s have demonstrated further improvements in fuel efficiencies, environmental control, waste gas, dust and byproduct recovery, as well as in the economics of the smelting process itself.



However, these have been superseded by the development of a new generation of flash smelting/flash converting by Kennecott and Outokumpu Oy at Garfield, Utah. The smelter at this site has been heralded as the "cleanest smelter in the world" and as such, one of the most significant innovations in extractive metallurgy in recent times (Emmons and Gabb, 1995). The new smelter and converter complex replaces an existing facility which was able to handle only 60% of the concentrates produced at the Bingham Canyon mine. To meet increasingly tough air quality regulations, the company was faced with a choice of investing $150 million in pollution control technology and being constrained by the existing smelter capacity, or investing $880 million on a new process. The new process increased the capacity of the smelter to handle 100% of the concentrates, thereby eliminating transportation and processing costs associated with the shipment of concentrates to Pacif Rim smelters, and enabling the plant to meet or exceed all existing and anticipated air quality regulations. It is anticipated that the new plant will reduce operating costs by 53% (Dimock, 1995). The principal features of the new complex are the replacement of traditional Pierce-Smith converters with a patented flash converter, the total enclosure of the converter, and the replacement of open-air ladle transfer of molten matte with a solid-state transfer. Molten matte is cooled with water into a granulated form prior to transfer to the converters, significantly reducing the release of sulphur dioxide and other gases in the transfer process. Although the cooling of the matte involves a loss of heat energy, "waste" heat is captured as steam and fed to a co-generation unit. The selection of flash converting enables a continuous high throughput of material and a much increased concentration of sulphur in the off-gas, greatly improving the efficiency of sulphur capture. In combination with the world’s largest double contact acid plant, annual average emissions of sulphur dioxide will be reduced from 3,600 pounds per hour to 200 per hour (Dimock, 1995; Chiaro 1994; Kosich,1995).

Lead

The traditional primary lead production route is based on sintering, reduction of the sinter in a blast-furnace, and refining of the bullion (which can be undertaken following pyrometallurgical or hydrometallurgical routes). Environmental controls to existing plants of this type require expensive addon controls, such as dust recycling systems, automatic pressure controls inside the sinter machine gas collecting system, electrostatic dust precipitation, de-drossing, higher shaft furnaces with new cooling systems, filter bags and hoods to collect waste gas.



Relatively recently, several new, more environmentally efficient smelting processes have been developed which apparently demonstrate energy, operating and investment cost savings over traditional smelting technology. These, however, are still at the demonstration stage. They all follow the principle, first developed in new copper production technology, of autogenous smelting where the natural heat generated by the oxidation of sulphide raw materials during roasting is used to smelt the charge in one single feed. This reduces overall toxic emissions through built-in dust and gas collection systems and by eliminating a heating stage (which produces off-gases) and indirect SO₂ emissions from the carbon fuel used in reverberatory smelters:

• The Kivcet Process uses an electric shaft furnace and collects vaporised metals in the off-gases in oxidic form. It contains built-in cleaning, ventilating and dust recovery systems at several points within the process. along the process;


• The Boliden (Top Blown Rotary Converter) Process involves dry-feeding ore to a furnace fuelled by injecting compressed air and oxygen through a lance. The furnace itself is located within a ventilated hood in which furnace tilting and pouring of lead bullion and slag. Slag granulation is undertaken, ventilated by bag filters if dry fumes are generated and by wet cleaning systems if moist gas is generated;


• The Outokumpu Process involves placing dry feed into a flash furnace via a special burner with oxygen injection under a ventilated conveying system. Waste gas from the flash furnace is neutralised by entering a special boiler and a hot electrostatic precipitator. After final cleaning in a wet electrostatic precipitator, gas is sent to an acid conversion plant. Dust is recycled in a closed conveying system to the flash furnace. Built-in filters, dust collectors and vacuum floor cleaning systems are located at different points in the process train;


• The QSL Process is currently scheduled for industrial operation at several European plants - works with moist feed. The charge is sent along ventilated conveyors to the furnace, which is equipped with one or two gas uptakes leaving the furnace at the oxidising zone or at the oxidising and reduction zone. Again, it contains built-in gas cleaning and dust collecting systems.

Zinc

One potential leap forward in the pyrometallurgical processing of zinc has been the Warner process, developed at the University of Birmingham (UK). In this process, zinc sulphide is converted directly to zinc metal using molten copper (which is thereby converted to copper sulphide). The zinc is vaporised and subsequently condensed under reduced pressure on a barrel condenser (Turner Jones and Warner, 1995). Energy given out during the conversion of the copper sulphide back to copper is transferred back to the process, and therefore large amounts of purchased energy are not required (Warner et al, 1994). However, the final zinc product requires refining to produce high grades.

Tin

Tin poses special environmental problems on account of its relatively high value. As a consequence of this, lower grades are processed involving many by-products such as copper and lead and impurities such as selenium, tellurium, antimony, bismuth and sulphur. Distinct smelting processes for different grades have been traditionally employed. Innovation has proceeded in two directions. First, it has aimed to remove and recover by-products more effectively. Second, it has aimed to use short rotary or electric furnaces depending on comparative fuel prices and series of filters and precipitators to collect toxic dust, particularly lead. Two new processes are currently under development: the Siromelt and TBRC smelting processes, which offer the advantage of consecutive treatments of metals and slag in the same vessel, avoiding the transfer of hot phases and reducing fugitive gas emissions.

Nickel

INCO’s development of oxygen flash smelting technology is an example of radical technical change necessitated by the exhaustion of possibilities for further efficiency improvements in conventional technologies. Until recently one of the world’s highest cost nickel producers, INCO was the greatest single source of environmental pollution in North America as a result of an aged and inefficient reverberatory furnace smelter at Sudbury, Ontario which emitted excessive volumes of SO₂. Having reached the limits of efficiency improvements and unable to meet increasingly stringent regulations as part of an intensive acid rain abatement programme by the Ontario Environment Ministry, INCO invested over C$3,000 million in research and development. The INCO oxygen flash smelter produces a concentrated SO₂ off-gas stream which can be efficiently captured and fixed as sulphuric acid. In addition the flash smelting process utilises the exothermic properties of sulphide ores and requires very little additional fuel. The process efficiencies stemming from the application of the technology have not only reduced SO₂ emissions at Sudbury by over 100,000 t/pa, but have helped transform the company into one of the world’s lowest cost producers (Warhurst, 1994).

Conclusions

As a means of summarising the environmental releases, effects and impacts from mining, mineral processing and extractive metallurgy, it is useful to take the perspective of the receiving (external) environment (Table 4).



Table 4 Physical, chemical and biological environmental effects of mining and milling

operations (after Eaton et al, 1994)

Increasingly, based on a more thorough understanding of these potential impacts, the pressure for environmental protection may outweigh the justification for exploiting particular mineral reserves (Hodges, 1995), and the industry as a whole, in partnership with the broadest possible spectrum of stakeholders, must strive towards ever higher levels of performance in control and management of the mining process throughout the life-cycle of an operation.

References

Almgren G, Almgren T and Kumar U. 1996. Just-in-time and right-in-space. Minerals Industry International, September, pp. 26-29.

Anonymous. 1991. (March). Phelps Dodge Corp. has announced the discovery of a substantial openpittable copper mineralization. Engineering and Mining Journal, March 1991 p. 71

Axelson O. and Sundell, L. 1978. Mining, Lung Cancer and Smoking, Scandinavian Journal of Work Environment and Health, 4, pp. 42-46.

Biswas, A.K., and W.G. Davenport. 1976. Extractive Metallurgy of Copper. Pergamon International Library, International Series on Materials Science and Technology, Vol. 20, Chapter 2.

Boldt J R. 1967. The Winning of Nickel. Toronto: Longmans.

Broad A. 1997. Arsenic down but not out. Metal Bulletin Monthly, No. 315, March 1997, pp. 54-57.

Chiaro P. 1994. Waste minimization and pollution prevention at Kennecott. In: Proc.

International Conference on Pollution Prevention in Mining and Mineral Processing (eds. K. Anderson and S. Purcell), pp. 100-107, Colorado School of Mines. CETEM. 1990. Analysis of Garimpeiro in Brazil, Unpublished research reports, Centro de Tecnologia Mineral, Rio de Janeiro, Brazil.

Clifford D. 1997. Mineral processing roundup. Mining Magazine, September 1997, pp. 176-190.

Coppin N J, Bryson F E and Brown C W. 1995. Pyrometallurgy and the environment: at what cost? In: Pyrometallurgy ’95, London: Institution of Mining and Metallurgy, pp 1-21.

Dimock R. 1995. Kennecott has modern mettle for mining. The Salt Lake Tribune, April 16 1995.

Dunster J and Dunster K. 1996. Dictionary of Natural Resource Management. Wallingford: CAB International, 363p.

Eaton P B, Gray A G, Johnson P W and Hundert E. 1994. State of the Environment in the Atlantic Region, Environment Canada.

Emmons C S and Gabb P J. 1995. Factors influencing a modernized copper smelter impurity control policy. In: Pyrometallurgy ’95, London: Institution of Mining and Metallurgy, pp. 125-142.

Godin E (ed). 1991. 1990 Canadian minerals yearbook - review and outlook. Energy, Mines and Resources Canada, Ottawa, Ontario.

Grant R M. 1994. Emerging developments in zinc extraction metallurgy. In: Metallurgical Processes for Early Twenty-First Century (ed H Y Sohn), The Minerals, Metals & Materials Society, pp. 125-149.

Hayes P C. 1993. Process Principles in Minerals and Materials Production, Hayes Publishing Company, Australia, 730p.

Hiskey J B. 1994. In-situ leaching recovery of copper: what’s next? In: Hydrometallurgy ’94, Chapman & Hall, pp 43-67.

Hodges C A. 1995. Mineral resources, environmental issues, and land use. Science, 268, pp 1305-

1312.

Kosich D. 1995. Kennecott’s vision may revolutionize smelter technology. Mining World News, 1

(3), pp .

Noranda Minerals Inc. 1990. Environmental Report. Noranda.

Nriagu J O and Pacyna J M. 1988. Quantitative assessment of worldwide contamination of air, water and

soils by trace metals. Nature, 333, pp. 134-139.

Rankin W J and Wright J K. 1992. Greenhouse strategies for the metallurgical industry. In: Minerals,

Metals and the Environment. Elsevier Applied Science, pp. 172-189.

Rogich D G. 1992. Trends in Material Use: Implications for Sustainable Development, Unpublished

paper, Division of Mineral Commodities, US Bureau of Mines, April 1992.

Sawyer J.W. 1977. Environmental Quality and the Extractive Industries: The Sulphate Issue. In:

Proceedings of the Economic Council, American Institute of Mining Engineers.

Thornton I. 1995. Metals in the Global Environment: Facts and Misconceptions. The International

Council on Metals and the Environment, Ottawa, Canada, 103p.

Turner Jones M D and Warner N A. 1995. Assessment of educed pressure condensation of zinc on a

barrel condenser. In: Pyrometallurgy ’95, London: Institution of Mining and Metallurgy, pp. 219-233.

US Congress, Office of Technology Assessment. 1988 (September). Copper: Technology and

Competitiveness. OTA-E-367. U.S. Government Printing Office. Washington DC.

USEPA. 1984. Overview of Solid Waste Generation, Management, and Chemical Characteristics.

Prepared for USEPA under Contract Nos. 68-03-3197, PN 3617-3 by PEI Associates, Inc.

USEPA. 1989e. Final Report: Copper Dump Leaching and Management Practices that Minimize the

Potential for Environmental Releases. Prepared by PEI Associates, Inc. (Hearn, R. and Hoye, R.)

under U.S. EPA Contract No. 68-02-3995.

van Zyl, D.J.A., I.P.G. Hutchison, and J.E. Kiel (editors). 1988. Introduction to Evaluation, Design and

Operation of Precious Metal Heap Leaching Projects. Littleton, CO: Society for Mining, Metallurgy,

and Exploration, Inc.

Warhurst A. 1991. Metals Biotechnology for Developing Countries and Case Studies from the Andean

Group, Chile and Canada. Resources Policy, March, pp. 54-68.

Warhurst. A. 1992. Environmental Management in Mining and Mineral Processing in Developing

Countries, Natural Resources Forum, 16(1), February.

Warhurst A. 1994. The limitations of environmental regulation in mining. In: Mining and the

Environment: International Perspectives on Public Policy (ed. R Eggert), Washington DC: Resources

for the Future. Warhurst A C and Bridge G. 1996. Improving environmental performance through innovation: recent trends in the mining industry. Minerals Engineering, 9(9).

Warner N A, Davies M W, Holdsworth M L and Turner Jones M D. 1994. Direct zinc smelting with

virtually zero gas emission. In: Metallurgical Processes for Early Twenty-First Century (ed H Y Sohn),

The Minerals, Metals & Materials Society, pp. 233-249.

Wills B A. 1997. Mineral Processing Technology, 6th edition, Butterworth-Heinemann, 486p.

WHO. 1985. Occupational Health of Miners in the Non-ferrous Metal Industry, UNEP Industry and

Environment, January-March. Worldwatch Institute. 1992. Figures based on production estimates in US Bureau of Mines, Mineral Commodity Summaries 1992, Washington, D.C.)

Young J.E. 1992. Mining the Earth, Worldwatch Institute, Paper 109, Washington, D.C., July.