Monday, September 20, 2010

Mineral Processing

1. Introduction
Hard rock mines almost always produce metallic ores (as opposed to mines that produce industrial minerals, salt, or coal). The metals are won from the ores of hard rock mines by procedures most often described as “mineral processing.”

The interface between the mine and mill (where the mineral processing is first performed) demands an exchange of knowledge and good communications. For this purpose, the miner must be versed in the basics of mineral processing. The text of this chapter is only a primer to a field of work that is far more extensive than can be adequately addressed in this handbook. Mineral processing procedures are complex and difficult to describe in simple terms. There are exceptions to be found to any general observation and the processes commonly employed are constantly changing due to technical advances and environmental concerns. In this chapter, license has been taken in some cases from true science and proper terminology to facilitate a basic understanding by the miner.

Mineral processing may be divided into three distinct phases.
• Comminution (crushing and grinding)
• Beneficiation (separation and concentration)
• Smelting and Refining

Comminution
Crushing was discussed in Chapter 25; comminution in this chapter deals with grinding the ore down to a fine particle size to facilitate subsequent processing.

Beneficiation
Beneficiation of hard rock ores may be separated into two fundamental processes – one for noble metals and the other for base metals.

Noble Metals
The first deals with the noble (“precious”) metals, such as gold and silver that are relatively inert and normally occur as fine particles of native metal in the ore. Noble metals are typically separated by gravity and/or leached from the ore (dissolved in a weak cyanide solution), and then precipitated or adsorbed and refined at the mine site to produce bars of bullion. These ingots are later further refined elsewhere to separate and purify the contained metals. Gold ores that may be treated in this manner are referred to as “free milling.”

Base Metals
The second deals with base metals, such as copper, lead, zinc, and nickel that usually occur in a chemical compound from which they cannot be separated in the mill. Instead, the mill separates the compound(s) from the ore by flotation to produce “concentrate(s),” which is the typical product from a base metal mine. The concentrates are then transported to a smelter and refinery that are usually a long distance away.

The concentrates often contain small quantities of precious metals (enhancing value) and may contain undesirable impurities, such as mercury or arsenic (reducing value). The premium paid by smelters for precious metal content is referred to as a “credit” and the reduction in payment because of impurities (or high moisture content) is most often called a “penalty.”

Base metal compounds are typically sulfides and most often occur in the ore along with other (unwanted) iron sulfide compounds, such as pyrite, marcasite, pyrrhotite, and arsenopyrite. Base metal ores from near the surface may have had the sulfide compounds converted to oxides by the force of nature. The oxide zone of a base metal deposit in hard rock is typically only in the order of 10 feet (3m) deep; however, there are cases where the oxidation penetrates to a depth of over 400 feet (120m). Oxide ores (particularly partially oxidized ores or “over-oxidized” ores that contain a native base metal) represent a challenge for the mill, which is normally designed to be efficient in dealing only with sulfide ores. On the other hand, oxide ores lend themselves to very economical extraction by direct leaching (for gold and silver) or SX-EW. At hard rock mines, the latter procedure is still mainly confined to copper ores. Operations that mine laterites (soft oxide ores of metals such as nickel and aluminum) are not considered to be hard rock mines.

Smelting and Refining
Since smelting and refining generally take place away from the mine site, these aspects of mineral processing are not pursued in this chapter.

Note
Hard rock miners use the same word (“mill”) to describe both the entire mineral processing plant (concentrator) and a machine used in this plant to pulverize the ore by tumbling it in a rotating drum. To avoid confusion in the text that follows, an adjective precedes the word mill (grinding mill, ball mill, or autogenous mill) when referring to the machine.

2. Rules of Thumb
General
• A concentrator (mill) requires up to 3 tons of water for each ton of ore processed. It is therefore important to operate with the with the maximum practical pulp density and minimum practical upward or horizontal movement. The basic philosophy requires movement over the shortest possible distances between processing units and makes use of gravity to save on power consumption. Source: Wayne Goul.
• In the arid climates, mills operate with less than one ton of new water for each ton of ore processed. The balance of the process water required is recovered from de-watering concentrate, thickening the tails, and re-circulation from tailing ponds. Source: Norman Weiss
• A mill at the mine (and related facilities) accounts for approximately 85% of the total electrical power consumption for an open pit operation, but only about 45% for a typical underground mine. Source: Alan O’Hara
• For a typical underground mine, the cost for electrical power for the mill (concentrator) will be approximately 35% of the total electrical power cost for the mine. Source: Fred Nabb
• A mill built entirely of second-hand equipment and controls may be constructed for half the cost of one built “all new” with state-of-the-art automated monitoring and controls. Source: Bruce Cunningham-Dunlop

Grinding
• Fine ore bins that provide feed to the grinding circuit should have a capacity equal to 30 hours of processing. Source: Northern Miner Press
• Grinding is a low-efficiency, power-intensive process and typically can account for up to 40% of the direct operating cost of the mineral processing plant. Source: Callow and Kenyen
• For purposes of design, it may be assumed that a ball mill will carry a 40% charge of steel balls; however, the drive should be designed for a charge of 45%. Source: Denver Equipment Company
• A grate (diaphragm) discharge ball mill will consume 15% more power than an overflow (open) discharge ball mill even though the grinding efficiencies are the same. Source: Lewis, Coburn, and Bhappu
• Other things being equal, the larger diameter the drum, the more efficient the grinding. However, this phenomenon stops when the diameter reaches 12.5 feet (3.8m). Thereafter, the efficiency bears no relation to diameter. Source: Callow and Kenyen
• The ball mill diameter should not exceed 100 times the diameter of the grinding media to avoid sloughing. Source: Bond and Myers
• In pebble mills, the individual pieces of media should be the same weight, not the same volume, as the optimum size of steel ball. Source: Bunting Crocker
• The power draft (draw) in a pebble mill can easily, quickly, and automatically be controlled to an extent that cannot be done on a ball mill. Source: Bunting Crocker
• The ratio of length to diameter of a rod mill should not exceed 1.4:1 and the maximum length of a rod (to avoid bending) is 20 feet. As a result, the largest rod mill manufactured measures fifteen feet diameter and is 21 feet in length. Source: Lewis, Coburn, and Bhappu
• For most applications, 70:1 is the maximum practical reduction factor (ratio) for a ball mill, but 60:1 represents typical design practice. Source: Jack de la Vergne
• Rubber liners in ball mills may have a service life of 2-3 times that of steel liners. Source: W. N. Wallinger
• The capacity of a mill with synthetic rubber liners is approximately 90% that of the same unit with steel liners. Source: Yanko Tirado
• The capacity of a grinding mill for a given product operating in open circuit is only 80% that of the same unit operating in closed circuit. Source: Lewis, Coburn and Bhappu.
• A dual drive (i.e. twin motors and pinions driving a single ring gear) may be more economical than a single drive when the grinding mill is designed to draw more than 6,000 HP (4.5 Mw). Source: Rowland and Kjos
• Geared drives are currently available up to 9,500HP. Source: Barrat and Pfiefer
• A direct drive ring motor (gearless drive) is the only option for an autogenous mill rated over 20,000 HP. Source: Mac Brodie

Gravity Separation
• For simple methods of gravity separation with water as the medium, a specific gravity differential of at least 1.5:1 between the minerals to be separated is desirable. Source: Fuerstenau, Peterson and Miller
• Common methods of gravity separation (jigs, tables and spirals) require close particle sizing and a difference in specific gravity differential of at least 2:1 between the minerals to be separated. Source: V. P. Kenyen
• For gravity separation to be possible, the ratio of the difference in density of the heavy mineral and the medium and the difference between the light mineral and the medium must be greater than 1.25. Source: Arthur Taggart
• Most all wet gravity separation equipment is sensitive to the presence of slimes (minus 400 mesh). Slimes in excess of 5% should be avoided. More than 10% causes serious separation problems. Source: Chris Mills

Flotation
• When designing the flotation circuit for a proposed mill, the scale-up factor for flotation retention times obtained from bench tests is approximately two. Source: Mining Chemicals Handbook (Cyanamid)
• To determine a preliminary water balance for a proposed flotation circuit, the pulp density may be assumed to be 30% solids (by weight). Source: Rex Bull
• As a rule, water-soluble collectors may be added anywhere in the circuit, but oily, insoluble promoters should always be added to the grind. Source: Keith Suttill
• For roasting to be exothermic to the extent that no fuel is required to sustain reaction, the flotation product must contain at least 17% sulfur. Therefore, the target is 18%. Source: Dickson and Reid

Filtration
• When designing the filters required for a proposed mill, the scale-up factor from bench tests is approximately 0.8. Source: Donald Dahlstrom
• When determining vacuum pumps for filter installations required for a proposed mill, the scaleup factor from bench tests is approximately 1.1. Source: Donald Dahlstrom

Concentrate
• The typical moisture content of concentrates shipped from the mine is often near 5%. If the moisture content is less than 4%, the potential for dust losses becomes significant. Source: Ken Kolthammer
• The moisture content of concentrate measured by a custom smelter will invariably be 1% higher than was correctly measured by the mine when it was shipped. Source: Edouardo Escala
• If the moisture content of the concentrate is above 8%, problems with sintering and combustion are usually avoided. Unfortunately, concentrates stored in a cold climate generally require maximum moisture content of 5% to avoid handling problems when frozen. Concentrate subject to both spontaneous combustion and a cold climate are usually dried to less than 4% and sometimes as dry as ½%. Source: Ken Kolthammer

3. Tricks of the Trade
• To calculate the area (acres) required to be cleared for construction of a mill and associated structures, simply divide the square root of the number for the daily tonnage of ore to be milled by 20. For example, a proposed 6,400 tpd mill will require 80/20 = 4 acres of land to be cleared. Source: Alan O’Hara
• To roughly calculate the quantity (cubic yards) of concrete required for a mill (not including a crusher house) to be built on a firm strata, multiply the square root of the number for the daily tonnage of ore to be milled by 60. For example, a proposed 1,600 tpd mill will require in the order of 40 x 60 = 2,400 cubic yards of concrete to be poured. Source: Jack de la Vergne
• To roughly calculate the power consumption in kWh/tonne of a proposed mill for a medium sized mine, add 15 to the number for the Bond work index (Wi) of the ore. For example, if Wi = 13, the power consumption will be approximately 28 kWh/tonne processed. Due to economy of scale, add only 10 to the Wi for a huge open pit copper mine, but add 20 for a small gold mine. (For more accuracy, refer to Chapter 23 – Electrical.) Source: Jack de la Vergne
• The Sixth-tenths Rule may be used to quickly obtain a rough estimate of the cost of building a proposed mill when the cost of a mill of different capacity processing the same mineral(s) is known and escalated to date. Reference: refer to Chapter 8 - Cost Estimating
• Jack’s Rule may be used to quickly obtain a rough estimate of the cost of building a proposed mill when the cost of a mill of similar capacity processing a different mineral(s) is known and escalated to date. Reference: refer to Chapter 8 - Cost Estimating
• The design of a new concentrator should consider provision to receive crushed ore for custom milling (or from a satellite operation) even though none is contemplated at the outset. Source: Jack de la Vergne
• To quickly estimate the size of a ball mill required for a proposed installation, first determine the drive HP and then select a corresponding mill size from manufacturer’s catalogues. The drive HP is calculated using Bond’s formula (see text). Then apply an assumed efficiency of 70-75%. Various Sources
• The drive efficiency of a smaller grinding mill with a typical induction motor drive can be quickly measured while it is operating by placing the prongs of a hand held ammeter around one of the feed cables. The value obtained is multiplied by √3 and again by the supply voltage to obtain power. The power value obtained (kW) divided by 0.7456 gives the actual horsepower draw to be compared with the nameplate horsepower of the motor drive to obtain the drive efficiency. Source: Unknown
• An operator can (and often does) raise the capacity of a grinding mill by increasing its speed (RPM) beyond the optimum, but the price paid is higher power consumption per ton processed. Source: Lewis, Coburn, and Bhappu
• An indicator of good ball wear is when the worn out balls discharging from the drum are 16 mm (5/8 inch) or smaller in size and are polygon shaped having as many as 12 surfaces that can be slightly concave. Source: Rowland and Kjos
• If a grinding mill is to be shut down for a long period of time, it should be emptied of the grinding medium and liners (if steel); otherwise, a permanent set can develop in journal (babbit) bearings. Source: Bob Dengler
• When the nature of the ore permits, it is more economical to grind all the ore just enough to partially liberate the mineral, then fine grind (regrind) only that portion of the ore that has reported to the rougher concentrate. Various Sources
• Hydro-cyclones give excellent overflows but poor underflows, resulting in very high circulating loads. The antidote is two stages of cyclones. Source: Bunting Crocker
• The relatively good copper recovery (95%) and clean concentrate (28.5 % Cu) at Granduc was partially due to the high temperature (30 degrees C.) of the flotation pulp. This was made possible because the cooling water from the power plant was used for process water. Source: J.J.M. Meyknecht
• Contrary to traditional beliefs that can be traced to an erroneous report on tests carried out by the USBM many years ago, lime does not depress gold in a flotation circuit when added in amounts up to 8 kg/tonne. Lime is less expensive than soda ash and it has a flocculating effect on iron oxide slimes, tending to depress them. Source: Fernando Benitez
• Gangue minerals should be scheduled along with metals. For example, excessive pyrite has the end effect of reducing the recoveries of lead and zinc into the concentrate, while silica impacts on the quality of zinc concentrate produced, downgrading its marketability. Source: Frank Kaeschager
• Too much froth in a secondary flotation circuit resulting from the re-circulation of mill water may be overcome with the addition of activated carbon. Source: Brooks and Barnett
• Hot vulcanizing in a rubber shop is the solution to the problem of separation of cold-bonded rubber linings in pipes, launders, pump boxes, etc. Source: W. N. Wallinger
• The problem of wood chips in return water can be overcome by installing a slightly submerged overflow in the thickener (drill some holes in the overflow launder). Source: G. Hawthorne
• In most cases, it is not economical to dry the concentrate more than 1% less than the moisture content that will incur an extra penalty at a custom smelter in the vicinity of the mine. Source: Rex Bull
• The recovery of gold from cyanide (pregnant leach) solution with activated carbon (CIP, CIL) is just as efficient and is typically installed for 65-85% of the capital cost and run at 80-90% of the operating cost incurred with a traditional zinc precipitation (Merrill-Crowe) installation. However, these economies do not hold true for a small mill or a larger mill treating rich gold ore. Source: John Wells
• The traditional Merrill-Crowe zinc precipitation process should be considered instead of CIP or CIL when the gold ore contains significant quantities of silver; otherwise the recovery of this metal will be halved. Source: Paul Chamberlin
• When “preg-robbing” (tendency to absorb gold from a cyanide solution) carbon or carbonaceous compounds occur in gold ore, recovery may be substantially reduced if the ore is treated as if it were free milling or simply refractory. In some cases, the “carbons” may be effectively removed (before leaching) by selective flotation. In other circumstances, roasting of a gold concentrate is the practical procedure. Various Sources
• There is often a security problem to transfer out gold bullion bars from a remote mine site in a developing country. A safe method is by helicopter, provided there is no fixed routine to the flight schedules. Source: Bruce Winfield
• Because of the security problems associated with gold bullion at a remote site in a developing country, consideration should be given (where the ore is amenable) to producing a flotation concentrate instead. The concentrate may be shipped out safely for further processing or sold directly to a custom smelter (who may pay a credit for a contained base metal that would otherwise be lost). Source: Jack de la Vergne

4. Grinding
Hard rock mine ores are invariably pulverized to a size small enough to liberate mineral particles from the barren rock (gangue). This comminution is ordinarily the first step that takes place within the mineral processing plant.

A number of grinding mill types are employed for hard rock mines. The classic grinding circuit consisted of a rod mill followed by a ball mill(s) in a two-stage circuit. This arrangement is still found at older installations and some newer ones that were built with used equipment.

A Semi-Autogenous Grinding (SAG) mill followed by a ball (or sometimes a pebble) mill(s) is the common arrangement found in modern plants of medium to large size in North America. Smaller mines often employ an extra stage of crushing to create product small enough to permit single stage grinding with a ball mill(s). For these reasons, only autogenous mills and ball mills are further described in this chapter.

Ball Mills
The ball mill remains the most widely used grinding unit at hard rock mines. The drum of a primary ball is typically cylindrical and of length equal to or up to 65% longer than its diameter. A number of even longer mills and conical mills have been manufactured in the past on the thesis that these designs better enable classification of the ground ore as it passes through the mill.

The grinding action is obtained by rotating the drum so that forged (or cast) manganese alloy steel balls (or cast iron slugs) are cascaded and tumbled with the ore. The ore is ground between balls and normally between balls and a steel liner. Over a period of time, the balls wear to a smaller diameter so that at any one time there is a gradation in the size. The average gradation is maintained by the regular addition of new (“green”) balls. In the past, steel balls had diameters ranging between two and three inches (depending on the drum diameter). Today, steel balls with four-inch diameters and more may be employed in larger diameter ball mills. The quantity (charge) of steel balls in the ball mill may range from 35 to 45% of the volume within. A mixture of crushed ore and water fills the space between and around the balls, such that the rotating drum is approximately half full. The pulp (crushed ore and water) in a ball mill is held near 75% solids (by weight).

The ball mill typically operates in closed circuit, meaning that a portion of its output (containing coarse ground ore) is recycled through the drum to be ground down to size. This recycling is a dynamic process in which pulp goes through the ball mill several times (on average). Between 2¼ and 2¾ times (225 - 275% re-circulation) is nominal; however, there are installations where the recirculation exceeds 500%. Separating the coarse fraction of the ground ore to be returned to the ball mill is normally accomplished in a hydrocyclone classifier. Rake classifiers and spiral classifiers are virtually obsolete, mostly due to the space required. 

Because it has no moving parts, the cyclone classifier requires little maintenance, but it consumes more power because the pulp must be pumped up to it at sufficient velocity to maintain 10 psi (70 kPa) or more of head at the entrance for proper performance. The nominal product from a ball mill is considered to be 80% -200 mesh. Larger particle size is termed a coarse grind while smaller sized product is referred to as a fine grind.

Case History
The world’s largest ball mills are the two installed at the 72,000 tpd Donde de Collahuasi mine in Chile. Each mill measures 22 feet in diameter by 36 feet in length and has a 12,500 HP shell drive. (The same mine has twin 10,500 HP SAG mills 33 feet in diameter by 15 feet long.)

Autogenous Mills
A few of the larger mines have been successful employing a FAG or simply AG mill (the larger chunks of crushed ore act as the grinding medium). This type of mill is very appealing (especially for a remote mine site), since it avoids the cost of purchasing, shipping, and handling grinding balls; however, it is only suitable for very hard ores with cubic cleavage. It is often extremely difficult to determine in advance whether a particular ore will work properly in a FAG mill.

A SAG mill can be described as a FAG mill that did not work properly with ore as the only grinding medium; therefore, steel balls were added. The ball charge is only about one-third of that required for a ball mill (usually 10-15% compared with approximately 40%).

The efficiency of a grinding mill depends on the weight of the grinding medium. This means that FAG and SAG (autogenous) mills are required to be of larger dimensions than a comparable ball mill because steel is 2½ to 3 times as heavy as the ore from a hard rock mine. However, the power consumption is similar, although some efficiency is lost in an autogenous mill because they typically require a grate (diaphragm) discharge to retain the coarse grinding medium while most ball mills have an open (overflow) discharge.

The drum diameter of an autogenous (FAG or SAG) mill manufactured in recent years on this side of the world is typically about equal to twice its length. For an autogenous mill to be most efficient, there is an optimum ore feed size that is related to the diameter of the mill and which may be determined by the following formula:
F = d80 feed (optimum) = 0.95D2/3
(Source: MacPherson and Turner)
Where 
d80 = size of opening (inches) through which 80% of the feed will pass.
D = the diameter inside the liners, measured in feet.
Example
1. Find the optimum feed size for a SAG mill 26 feet in diameter.
2. Find the open-side (o/s) setting of an underground crusher to obtain this feed on surface, assuming an attrition of ½ inch in the transport and storage of ore between the underground crusher and the SAG mill.
Facts: 
1. D = 26 feet
2. Attrition = ½ inch
3. The product of this crusher is ½ inch less than the open side setting
Solutions: 
Optimum d80 Feed Size = 0.95 x 262/3 = 8½ inches
Open-side setting (o/s) = 8½ +½ +½ = 9½ inches

Case History
For many years, the world’s largest autogenous mill was an Allis-Chalmers 12,000 HP FAG mill 36 feet in diameter by 15 feet long, installed at Hibbing, Minnesota. Recently, a 40-foot diameter SAG mill was reported to have been designed and manufactured.

Grinding Mills (Autogenous Mills and Ball Mills)
Critical Speed and Optimum Speed
The critical speed, Cs (measured in RPM) is when the centrifugal force on the grinding mill charge is equal to the force of gravity so that the charge will not tumble as the drum rotates. Cs is calculated using the following formula.
Cs = 76.63√D
Where :
D = the diameter inside the liners, measured in feet.
Optimum crushing efficiency is obtained when a grinding mill is run at a particular fraction of critical speed. It is often reported in the literature that the optimum speed is near 75% of critical. This is true of a ball mill that is 10 feet (3m) diameter, but the optimum speed is greater for a smaller diameter ball mill (80% for a 3-foot diameter ball mill). Optimum speed is typically less than 75% for one of larger diameter (as low as 65% for a 20-foot diameter ball mill).

Bond’s Law
During the 1940’s, Fred Bond (largely in association with W. L. Maxon) developed a system for comparing ore grindability in terms of weight passing a specific mesh size per revolution of the grinding mill. Since that time, others have developed similar analyses, but the original system prevails today for grinding mills (and may also be used for crushers).

Bond’s formula is conveniently expressed as follows.
W =Wi (10/√P -10/√F)
W = work (kWh/short ton ore)
P= size in microns (μ) through which 80% of the product passes (P80)
Wi = work index
F= size in microns (μ) through which 80% of the feed passes (F80)
Bond’s formula contains a mixture of metric and Imperial units. To convert to all metric, the denominators (10) are simply changed to 11 to obtain the result in kWh/metric ton (tonne).
W =Wi (11/√P -11/√F)
Some metallurgists add modification factors to the Bond formula in comprehensive calculations to obtain greater accuracy. Table 26-1 provides typical work indices for some common rocks and minerals. For purposes of designing a proposed grinding mill, the work index of the ore to be treated is obtained from
laboratory test reports.
Table 26-1 Bond Work Index for Rocks and Minerals


Table 26-2 provides particle sizes in microns (μ) required for use in the Bond formula.
Table 26-2 Feed and Product Sizes in Microns (μ)


(1μ = 1 x 10-6m)
Bond’s Law Example
Calculate the reduction ratio and estimate the power consumption of a ball mill, using the Bond formula.
Facts: 
1. F, the feed is from a cone crusher with a 5/8-inch open side setting
2. P, the product desired is 70% passing a Tyler 65 mesh screen (P70 = 65 mesh)
3. Wi, the work index of the ore to be ground is 15
Solution: 
1. From the Feed and Product Size Table, the feed, F80 = 15,000μ for a 5/8 inch openside setting.
2. From the Feed and Product Size Table, a product, P80 = 210μ for 65 mesh.
3. The desired product size, P70 = 210 x (80/70)2 = 274μ for 65 mesh.
4. Reduction ratio = F/P =15,000/274 = 55 (55:1).
5. Power, W = Wi (10/√P -10/√F) = 150(1/√274 -1/√15,000) = 7.8 (7.8 kWh/short ton).

Controls
The efficiency of a grinding mill is dependent not only on the optimum RPM of the drum, but also the ball charge and the rate and blend of feed. These multiple variables make it difficult even for seasoned operators to manually maintain optimum efficiency in the grinding circuit. When the efficiency of a dynamic process is dependent on multiple variables, computerized controls and simulation modeling are advantageous. Computers have controlled grinding circuits in some mills for over 20 years. These controls are credited with increasing the efficiency of grinding circuits by 5% and more.

Shutdown and Salvage
A large value of gold may be recovered from a grinding mill that has operated for many years in a mine containing gold in the ore. Ores containing gold also contain minute amounts of mercury, chlorine, silver chloride, etc. that are released in the milling process. Gold combines with the other materials and collects as a crude amalgam in every crevice and surface in the grinding mill (not subject to direct abrasion). The amalgam is invisible because it is the same color as steel; however, the amalgam is softer and can be readily identified and removed with a hammer and cold chisel.

After removal, mercury, soda ash, and lead nitrate are added to the amalgam, which is then ground and pressed to remove excess mercury. The compressed material may be then put in a laboratory retort to distill off (and recover) mercury and leave behind a dirty sponge of gold to be washed and refined.

5. Beneficiation
At hard rock mines, numerous methods are employed to separate and concentrate the ore (beneficiation). The most common methods are flotation, leaching, gravity separation, and solvent extraction. The most widely employed is flotation. For ores of many metals, flotation is used exclusively, for some it is employed in conjunction with another method, most often gravity separation. The most common application for precious metal ores is leaching. Solvent extraction is usually peculiar to uranium ores and copper oxide ores. Consequently, this chapter further describes only flotation, leaching of precious metals, and gravity separation.

Flotation
Due to surface tension, a tiny flake of gold can float in a glass of water despite the fact that gold is almost twenty times as heavy as water. If the water is stirred vigorously with a spoon after a few drops each of pine oil and creosote (reagents) are added, a minute (<75μ) gold particle will usually rise and float in the thin surface foam (froth). This occurs because the reagents have made the tiny gold particle water-repellant and bubble attractive. Commercial flotation exploits and enhances this phenomenon to float (and skim off) fine particles of gold in unusual circumstances, such as when fine gold is contained in a cyanogenic (over-oxidized) gangue or freely associated with other minerals (to be recovered by the same process).

Base metal minerals are much more amenable than gold to this same procedure and they constitute by far the most common application for recovery by flotation. These minerals (usually sulfides) occur as particles (separated by the grinding process) that are recovered by the same process of flotation from “pulp” (a mixture of finely ground ore and water).

In some cases, flotation is used to remove an unwanted mineral component from the pulp. Reagents usually work better in an alkaline pulp; therefore, a pH regulator (usually lime) is typically added before the reagents are meted out. The reagents used are mainly the following proprietary chemical compounds.
• A frother (pine oil is a simple frothing agent with some collecting powers)
• A collector and promoter (combined in one agent or provided separately)
• A modifier (which may have one of any number of functions)

Air is a necessary component of the flotation procedure. Usually, slightly compressed air is blown into the cells providing agitation and a thicker froth (bigger bubbles). A few mills employ simple mechanical agitation (analogous to stirring with a spoon), because too much air is deemed detrimental to flotation efficiency in their particular case.

Selective flotation refers to floating one mineral while leaving behind (depressing) another. Differential flotation refers to selective flotation of different economic minerals in succession from a poly-metallic ore. In an elementary flotation circuit, the pulp is “conditioned” with reagents before flowing continuously through one bank of cells (tier of connected, open rectangular tanks) and then to another bank. The float is continuously skimmed off in each cell. The skimmed float from the first stage of cells (roughers) is typically combined with float from a second stage (scavengers) and directed to a third stage (cleaners) where it is re-floated to produce a slurry of nearly pure mineral (concentrate or “cons”). The slurry is subsequently de-watered and further dried to make the final concentrate that is then suitable as smelter feed.

The barren pulp (“tails”) from which the mineral has been won is partially de-watered and piped from the mill for further de-watering and permanent disposal in a tailings impoundment or some used as backfill in empty underground stopes (as a component of paste fill or hydraulic fill).

In new mills, the flotation cells are larger (the silting problem having been solved). Often, the traditional rectangular tank cells in the cleaning phase are replaced with erect cylindrical tanks (column flotation). These advances save space, facilitate monitoring, and reduce maintenance costs; however, there is a practical limit to the size of cells to prevent short-circuit and maintain flexibility.

The larger the mill capacity, the larger the cells.
As is the case with grinding, modern flotation circuits are equipped with computerized controls to regulate and monitor variables, such as grind, pH, reagent feed, conditioning residence time, cell retention time, and pulp density.

Leaching of Precious Metals
Mines most often separate and recover noble metals, such as gold (and the silver that is normally found with the gold) by gravity and/or cyanide leaching. Gravity concentration is used almost exclusively when placer mining. Some hard rock mine mills use gravity as a preliminary step in the recovery procedure, but for many it is either inefficient (because of minute gold particle size) or not desired (because it is perceived to facilitate theft).

In the normal (free milling) process, the ground ore is treated (leached) with a dilute alkaline (high pH) solution of sodium cyanide (NaCN). The solution is made alkaline with addition of generous quantities of hydrated lime (quick lime). The cyanide reacts chemically with the gold to form a soluble compound, Au(CN)2 that is dissolved. The solution containing gold is called pregnant or simply “preg.”

Subsequently, the pregnant solution may be clarified and the gold directly precipitated with zinc powder (traditional Merrill-Crowe process). The precipitate is then compressed (pressed) and then the compacted precipitate is smelted on-site to produce bars of bullion.

More commonly, the Au(CN)2 is adsorbed by pellets of activated carbon that are later separated by screening. The pellets are then heated (230-266 degrees F) and flushed with an alkaline cyanide solution to dissolve (strip) the gold. The dissolved gold is typically recovered from the resulting rich solution by electrolysis (“electrowinning”).

The cyanide solution from which the gold was precipitated or the loaded carbon was separated is termed barren and is recycled. So are stripped carbon pellets (which are first re-activated in a kiln). When the loading of the carbon takes place subsequently and separately from leaching, the process is called CIP. Loading coincident with leaching (short circuit) is called CIL.

Some gold ores are not free milling. In rare instances, the gold in the ore is found to be chemically combined (tellurides) and more commonly found locked into sulfide crystals (“refractory”). Refractory ores are amenable to flotation and the concentrate obtained then subjected to a process (roasting, biological leaching or auto-claving) that reduces the crystalline sulfides to oxides that may subsequently be efficiently leached with cyanide.

Cyanide has been the leaching agent of choice for a century despite the fact that it is expensive and lethal if swallowed. There are substitutes, but none have thus far proven effective and practical. In North America, only one recorded accidental fatality has occurred from cyanide poisoning in the mining industry in the whole of the 20th century. Nevertheless, cyanide is now outlawed or otherwise forbidden in some jurisdictions and this may be the beginning of an unfortunate trend.

Gravity Separation
Gravity separation is the oldest and apparently the simplest means of concentrating ores. Gravity separation is often defined as the separation of coarse particles of metal or mineral; however, this is not always the case. Artificial gravity (centrifugal force) can enable the separation of fine particles in a centrifuge or cyclone.

To comprehend gravity separation, it is necessary to understand that a difference in weight alone (differential specific gravity) is not the only consideration for separation in a dynamic process. Particle size plays an equally important role, as does the medium in which the particles are suspended (air, water, brine, or heavy media). A dense particle (high specific gravity) of tiny size may have the same amount of movement in a fluid media as a different larger particle that is lighter (lower specific gravity). It follows that for a difference in specific gravity to be effective, all the particles should be nearly the same size. Conversely, artificial gravity may be employed to separate particles of similar specific gravity but different particle size (“classification”). This is the operating principle of the hydrocyclone classifier that re-circulates the coarse fraction of ground pulp back into a ball mill, as previously described. Besides centrifuges and cyclones, all sorts of machines are used for gravity separation, including sluices, jigs, cones, shaking tables, spirals, and heavy media separation.