Gravity concentration is one of the oldest approaches in the history of minerals extraction — ancient prospectors understood intuitively that gold settles faster than sand and that a well-designed sluice could capture it. What has changed dramatically in modern times is the engineering precision and scale at which density-based separation is practiced, and the range of minerals to which it is applied. Today, gravity concentration in mineral processing recovers iron ore by the billion tonne, processes beach sand mineral deposits for titanium and zirconium, cleans billions of tonnes of coal annually, and captures free gold in grinding circuits worldwide. It accomplishes all of this with lower reagent costs, lower environmental impact, and greater operational simplicity than flotation. This guide provides a thorough treatment of all the major gravity concentration technologies — from the physics of density-based separation and the design of spiral concentrators, through shaking tables, jigs, centrifugal concentrators, and sluices, to circuit design principles and gold recovery applications.
What is Gravity Concentration? Principles and Applicability
Gravity concentration is the separation of minerals based on differences in particle density. The density difference causes particles to move at different rates and in different directions within a fluid or flowing-film environment, allowing them to be physically separated into products of different mineral compositions. Water is the carrier fluid in almost all industrial applications, though air is used in certain dry-processing situations where water is unavailable or where product quality requirements favor dry handling.
The Concentration Criterion
The feasibility of gravity separation for a given ore is often assessed using the concentration criterion Δρ:
Δρ = (ρh − ρf) / (ρl − ρf)
where ρh is the density of the heavy mineral, ρl is the density of the light mineral or gangue, and ρf is the density of the fluid medium. When Δρ exceeds 2.5, gravity separation is relatively easy and can be performed effectively on particles as fine as 75 μm. Values between 1.75 and 2.5 indicate that separation is possible but requires closer size control, with efficient separation typically requiring particles coarser than about 150 μm. Below 1.75, gravity separation becomes progressively more difficult and eventually impractical below a criterion of about 1.25.
To illustrate: gold (s.g. 19.3) separating from quartz (s.g. 2.65) in water gives a concentration criterion of 11.1 — which is why simple panning works so well. Cassiterite (s.g. 7.0) from quartz gives a criterion of about 4.3, making spiral and table concentration very effective. Ilmenite (s.g. 4.7) from quartz gives about 3.5, while a separation of fluorite (s.g. 3.2) from quartz gives about 1.6 — possible but requiring careful size classification and controlled operating conditions.
The Role of Particle Size
Particle size profoundly affects the efficiency of gravity concentration. As particle size decreases, the ratio of surface forces to mass forces increases, the settling velocities of all particles become very similar regardless of density, and Stokes’ law governs behavior. Below about 50 μm the differences in settling velocity between heavy and light minerals become so small that conventional gravity methods struggle to achieve useful separation. This sets a practical lower limit on the particle size that can be treated by conventional gravitational methods. Enhanced gravity concentrators, which apply centrifugal rather than purely gravitational acceleration, extend the range of effective gravity concentration to finer particle sizes, but even these devices have practical limits.
The size effect also means that feed preparation is critical for efficient gravity concentration. All gravity separators benefit from feed that has been classified into narrow size fractions before processing, because size variation within the feed creates conflicting separation forces. Large light particles and small dense particles have similar settling velocities — they cannot be separated by their settling behavior. Close sizing eliminates or reduces this ambiguity, greatly improving separation efficiency. In practice, this is achieved by screening coarser feeds and by using hydraulic classifiers (multi-spigot hydrosizers) for finer feeds before shaking tables or fine spirals.
Advantages and Limitations
Gravity concentration offers compelling advantages: it requires no chemical reagents (reducing both costs and environmental footprint), operates at ambient temperature and pressure, involves mechanically simple equipment with relatively low capital and operating costs, and produces tailings that are generally easier to dispose of than tailings from flotation circuits laden with flotation reagents. These advantages explain the renewed interest in gravity methods in recent decades as environmental regulations have tightened and reagent costs have risen.
The primary limitations are the particle-size dependency discussed above, sensitivity to feed slurry density variations, and difficulty in separating minerals of similar density. Slimes — particles below about 10–20 μm — are particularly problematic because they increase slurry viscosity, reduce the sharpness of separation, and tend to coat coarser particles, interfering with surface-based separation mechanisms. De-sliming prior to gravity concentration, typically using hydrocyclones, is therefore standard practice, though at the cost of fine-particle losses.
Spiral Concentrators: Design, Operation, and Circuit Configuration
The spiral concentrator is the workhorse of large-scale gravity concentration. Introduced commercially in 1943 as the Humphreys spiral, it processes feed in the size range roughly 3 mm to 75 μm and handles capacities up to 3 t/h per single spiral turn on mineral sands applications, making it highly efficient on a capital and operating cost basis. Spirals are used for iron ore, mineral sands (ilmenite, rutile, zircon, monazite), chromite, fine coal, tin, tungsten, and gold concentration.
How a Spiral Concentrator Works
A spiral concentrator consists of a helical trough — typically with a modified semi-circular cross section — wound around a central support column for five to eight turns. Feed slurry, at 15–45% solids by weight, is introduced at the top. As the slurry flows downward under gravity along the helical path, several forces act simultaneously: the primary flow along the channel axis, a secondary cross-channel flow driven by centrifugal effects (flowing outward at the top of the stream and inward near the channel floor), differential settling rates of particles of different density and size, and interstitial trickling of finer particles through the bed.
The combined effect is stratification across the width of the channel: dense particles migrate toward the inner edge of the spiral where velocities are lowest, while lighter particles are carried to the outer edge in the faster-moving outer current. This stratification builds progressively over the spiral turns, creating a visible band structure with dark, dense mineral at the inner edge and light gangue at the outer edge. Adjustable splitters at the base of the spiral — or at intermediate take-off ports positioned at intervals along the spiral — divide the stream into heavy product (concentrate), middlings, and light product (tailings).
Wash water, introduced along the inner edge of the channel, flows outward across the concentrate band to help flush entrained light particles back into the tailings stream and sharpen the separation. Some modern spiral designs operate without wash water — so-called wash-waterless spirals — which simplifies operation and reduces water requirements at some cost to separation efficiency.
Key Operating Variables
The most important operating variables for a spiral concentrator are feed pulp density, feed rate, splitter position, and wash water rate. Feed pulp density significantly affects the separation: too dilute and the benefits of hindered settling are lost, producing poor stratification; too dense and viscous effects dominate, also reducing efficiency. Optimal feed density typically falls in the range 25–35% solids by weight for mineral sands applications, though this varies by ore type. Feed rate directly affects the residence time and the depth of the flowing film — higher feed rates reduce both and shift the size-density distribution of the concentrate. Splitter position is the primary control for adjusting the grade-recovery balance between concentrate and tailings.
Modern Spiral Designs
Modern spiral design has diverged considerably from the original Humphreys design. Double-start spirals — two spiral channels wound around a shared central column — have effectively doubled the capacity per unit of floor space without significantly affecting separation performance, and are now the standard configuration in most large-scale operations. Spiral slopes (helix angles) are matched to the specific gravity of separation: shallow pitches for separations with small density differences (such as coal from shale), steeper pitches for large density differences (such as zircon from staurolite). Typically, roughing spirals have five to eight turns, while cleaning spirals may have only three turns to allow a finer product split. Capacity ranges from about 1 t/h on shallow-pitch coal spirals to about 3 t/h on steeper mineral-sand spirals per single channel.
Circuit Configurations
Single-stage spiral circuits are rarely sufficient for final concentrate production. Multistage rougher-cleaner and rougher-cleaner-recleaner arrangements are standard, and the introduction of compound spirals in which a second-stage separation is built into the same column has reduced floor space and capital cost compared to two physically separate stages. Analytical work by Luttrell (2002) demonstrated that a rougher-cleaner spiral circuit with middlings recycle achieves a sharpness of separation that is 1.33 times greater than a single spiral unit at the 50% recovery point — a significant efficiency gain obtained without any equipment addition. This same principle underlies the design of multi-stage spiral circuits in iron ore processing, where rougher, scavenger, and cleaner spirals operate in combination to achieve high iron recovery at grades suitable for blast furnace feed.
Shaking Tables: Wilfley Tables and How They Work
The shaking table — sometimes called a Wilfley table after one of its earliest commercial designs — is the most metallurgically precise gravity concentrator available. It is used for the final upgrading of gravity concentrates, for the treatment of small difficult-to-process streams, and for laboratory and pilot-scale testwork. Its ability to produce visible fan-shaped separation patterns that allow the operator to see the separation in real time makes it invaluable for troubleshooting and metallurgical testing.
Table Construction and Separation Mechanism
A shaking table consists of a slightly inclined flat deck — typically 1.5–2.5 m wide and 3–4.5 m long — mounted on a vibrating mechanism that produces a characteristic asymmetric reciprocating motion: a slow forward stroke followed by a rapid return. This motion causes particles to be conveyed forward along the long axis of the table. Simultaneously, wash water flows laterally across the deck, perpendicular to the direction of motion, and the deck tilts slightly downward in the direction of wash water flow. Particles are thus simultaneously driven forward by the table motion and sideways by the wash water, so they travel diagonally across the deck, and the angle of their trajectory depends on the balance between these two forces acting on each particle.
Since the wash water force on a particle depends on its size (larger particles experience more drag) and its density (heavier particles resist lateral displacement more effectively), the diagonal trajectories fan out across the deck. Small, dense particles travel nearly parallel to the long axis (high on the table) toward the concentrate launder. Large, light particles are swept more strongly sideways by the wash water and travel toward the tailings launder. Middlings fan out in between. This visible fan pattern can be directly observed and the splitters adjusted to cut the product streams at any desired point on the grade-recovery curve.
The Role of Riffles
Along most of its length, the deck surface carries tapered wooden or composite riffles running parallel to the long axis of the table, with maximum height on the feed side and tapering to zero near the opposite side. The riffles create protected pockets behind them where multilayer beds of particles accumulate and stratify: finest and densest at the bottom, coarsest and lightest at the top. The taper means that as material moves toward the low-riffle side, progressively finer and higher-density material is exposed to the flowing film of wash water and carried forward. The final clean-up occurs on the unriffled section of the deck, where the feed stream is a single or double layer of particles and the separation is at its sharpest.
Feed Preparation and Capacity
Because the shaking table operates on such a precise flowing-film mechanism, it is highly sensitive to the size distribution of the feed. The standard practice is to classify table feed in multi-spigot hydrosizers, feeding each narrow size fraction to a separate set of tables. Feeds classified as sand tables handle the fraction from about 3 mm to 100 μm; finer fractions (slimes) are fed to tables fitted with planes rather than riffles. Capacity ranges from about 0.5–2 t/h of ore per deck, depending on feed size; on coal feeds at up to 15 mm, capacities of 12–15 t/h per deck are achievable. Double- and triple-deck tables increase area efficiency but at the cost of some individual deck control flexibility.
Operating Variables
The primary operating variables on a shaking table are deck slope (adjusted across both the lateral and longitudinal directions), wash water rate, feed pulp density, stroke length, and stroke frequency. The line of separation is clearly visible on the table surface, so experienced operators can make adjustments based on direct observation. For fine feeds a shorter, faster stroke is appropriate; for coarser feeds a longer, slower stroke. The correct deck slope is especially important: too steep, and concentrate grade suffers; too flat, and recovery drops as dense particles are carried into the tailings by the wash water. Feed pulp density of 20–25% solids by weight is standard for metallic ore applications.
Jigs: Pulsating Jigs and Applications
Jigging is one of the oldest mineral concentration techniques, described in Agricola’s De Re Metallica (1556), yet modern jigs incorporating electronic control, improved mechanics, and optimized geometry remain highly competitive for coarse separation applications. Jigs are particularly strong performers in coal washing, iron ore beneficiation, and the recovery of cassiterite, tungsten, gold, and barytes from coarse feeds.
The Jigging Mechanism
A jig consists of an open tank filled with water, with a horizontal screen at the top supporting a bed of coarse dense particles called ragging. Feed slurry flows across the ragging while a pulsating water current — produced by a reciprocating piston, diaphragm, or compressed air — causes the bed to alternately dilate (expand) and compact. During the dilation phase of each pulse, the bed opens and particles are classified by their settling velocities in the upward water current. During the suction phase, the bed compacts and smaller particles trickle downward through the interstices of the ragging, a process called consolidation trickling.
The net effect over many cycles is stratification by density: dense particles migrate downward through the ragging and ultimately pass the screen to be discharged as the heavy product (concentrate) via the hutch, while light particles are swept horizontally across the bed surface by the feed flow and discharge as the light product (tailings). The jigging action is most effective for particles coarser than about 150 μm; below this size, the short settling time per pulse and the increasing influence of surface forces make efficient stratification progressively more difficult.
Types of Jigs
The Harz jig, with a plunger-type piston operating in a separate side compartment, was the classic design for metallic mineral applications and is still found in older plants treating cassiterite, tungsten, or gold. The Denver mineral jig, widely used in grinding circuits to remove liberated heavy minerals before they can be overground, uses a rotary water valve synchronized with the piston to give precise control of the hutch water addition. The circular radial jig, developed in 1970 and used extensively on tin dredges in Southeast Asia, uses a trapezoidal cross-section arranged in sectors of a circle, so the cross-flow velocity decreases radially as the feed moves outward, maintaining uniform separation conditions across the full area of the bed.
The InLine Pressure Jig (IPJ), developed by Gekko Systems, represents a significant innovation: it is fully encapsulated and operates under pressure, completely filled with slurry. This design eliminates the conventional feed-to-screen freeboard, reduces water requirements, and allows the unit to be installed directly in grinding circuit pipelines. IPJs are used widely for free gold recovery, sulfide recovery, and cassiterite concentration in grinding circuits.
In the coal industry, the dominant jig types are air-pulsated. The Baum jig, which has been in service for nearly 100 years with continuous design refinements, uses a large air chamber on one side of the machine to produce pulsation across the full width of the jig bed. For very large tonnages (up to 1,000 t/h) of a wide size range, the Baum jig remains competitive. The Batac jig improves on the Baum design by placing multiple air chambers extending the full width of the jig from below, providing more uniform pulsation across the entire bed and permitting more precise control of stroke characteristics through electronically governed air valves.
JigScan Control
Modern jigs increasingly incorporate automated control systems. The JigScan system, developed at the Julius Kruttschnitt Mineral Research Centre, measures bed conditions and pulse velocity many times per pulse cycle using pressure sensors and nucleonic gauges. The data provide detailed information about bed behavior in real time and allow the control system to detect and correct abnormal conditions. Reported improvements in coal yields of greater than 2% over uncontrolled operation demonstrate that automated jig control offers measurable and commercially significant benefits.
Centrifugal Concentrators: Knelson and Falcon Concentrators
Conventional gravitational methods lose efficiency below about 50–75 μm because the settling-velocity differences between dense and light particles become very small. Centrifugal concentrators overcome this limitation by substituting a centrifugal acceleration of tens to hundreds of times gravity for the normal gravitational field, restoring the relative settling velocity differences and enabling efficient concentration of fine particles that conventional methods cannot treat.
Knelson Concentrator
The Knelson concentrator, developed by Byron Knelson in the 1970s and first installed commercially in 1987, is the most widely used semi-continuous centrifugal concentrator in the gold mining industry. It consists of a rotating conical bowl driven at speeds generating centrifugal accelerations of up to 60G. Feed slurry is introduced through a central stationary feed tube into the bottom of the rotating cone, where it is immediately subjected to the centrifugal field. The slurry moves up the cone wall under centrifugal force, filling a series of concentric riffles (rings) machined into the cone interior.
Each riffle acts as a small compartment in which a fluidized bed of dense particles accumulates. Fluidization water is injected through holes in each riffle from behind, keeping the bed in a fluidized, dynamic state that allows ongoing exchange of particles between the bed and the flowing slurry. Dense particles such as gold, platinum, or sulfide minerals are captured and retained in the riffles while light gangue particles are continuously flushed outward over the tops of the riffles and discharged. At the end of the concentration cycle — typically every 30–90 minutes in automated operation — feed is stopped, the bowl is slowed, and the accumulated concentrate is flushed out with water. The entire discharge cycle takes less than two minutes, minimizing downtime.
Knelson concentrators treat feed particles from 10 μm up to 6 mm at capacities ranging from a few kilograms per hour in laboratory models to 150 t/h in the largest industrial units. They are most effective where the dense component constitutes less than about 500 g/t (0.05%) of the feed — the target application being free gold recovery from grinding circuits. The unit has gone through multiple design generations from the original Manual Discharge (MD) through Center Discharge (CD), Extended Duty (XD), and the Quantum Series, each improving concentrate grade, recovery, or operational security.
Falcon Concentrator
The Falcon concentrator, first installed commercially in 1986, operates on a similar rotating bowl principle but uses a different fluidization approach. Feed slurry enters the bottom of a spinning bowl-shaped cone and flows up the sloped walls under centrifugal force. Unlike the Knelson, the Falcon operates without internal fluidization water in its Semi-Batch (SB) model: instead, the concentrate bed is fluidized by the process slurry pressure itself (backpressure process water), so the unit operates at higher rotational speeds than the Knelson. The higher centrifugal acceleration — up to 300G in some models — makes the Falcon particularly effective at capturing fine or flaky gold that may be lost in Knelson units, and case studies have shown that Falcon units can outperform Knelson concentrators on flotation feeds and concentrates where fine or irregularly shaped gold is present. Continuous Falcon models (Falcon C) are available for applications requiring uninterrupted operation.
Kelsey Centrifugal Jig
For coarser centrifugal gravity concentration, the Kelsey Centrifugal Jig (KCJ) places a conventional jig bed in a spinning centrifuge. The centrifugal acceleration amplifies the jigging action, extending the effective size range of jigging to finer particles. The KCJ handles feeds up to about 100 t/h in its largest version and has been applied to tin, gold, chromite, and tantalum recovery. It is particularly useful where the feed contains both coarse and fine particles that overlap in settling behavior at normal gravity.
Multi-Gravity Separator
The Multi-Gravity Separator (MGS), developed by Richard Mozley, can be conceptualized as a shaking table rolled into a rotating drum. The drum rotation generates centrifugal accelerations of 5–15G at the drum surface, while a superimposed sinusoidal shaking motion mimics the reciprocating motion of a conventional shaking table. A scraper within each drum conveys the settled dense fraction up the drum slope, where it is countercurrently washed before discharge as concentrate. The MGS has demonstrated significant improvements in fine cassiterite and chromite concentrate grade and is particularly valuable for cleaning fine gravity concentrates that conventional tables cannot treat efficiently.
Sluices and Riffle Boxes
Sluices and riffle boxes are the simplest gravity concentration devices and historically the most widespread, used from ancient Roman mining to the gold rushes of the nineteenth century and continuing in modern artisanal and small-scale mining today. In the mineral processing industry they are used for high-capacity roughing of beach sand mineral deposits and for preparatory concentration of coarse alluvial or placer materials before final upgrading on spirals or tables.
Pinched Sluices
The pinched sluice is an inclined launder approximately 1 m long that narrows from about 200 mm width at the feed end to about 25 mm at the discharge. Feed slurry at 50–65% solids by weight is introduced with minimal turbulence. As the slurry descends the sluice, it stratifies: dense particles sink to the bottom of the flowing film while light particles remain near the surface and travel faster. At the discharge end, the stratified layers are split by a splitter or tray arrangement to give a dense fraction and a light fraction. The separation efficiency of a simple pinched sluice is relatively low, but its extreme simplicity and minimal maintenance requirements make it attractive as a roughing device ahead of more efficient equipment.
Reichert Cone
The Reichert cone, designed in the early 1960s primarily for titanium-bearing beach sands, applies the pinched sluice principle at much larger scale using a series of conical surfaces arranged vertically in a single machine. Feed slurry is distributed around the outer edge of each cone and flows inward toward the center; a slot near the center removes the dense fraction while the light fraction flows over the slot. Multiple cone stages within a single machine achieve the separation efficiency that a single stage cannot, though the separation efficiency per stage remains lower than that of spirals. The Reichert cone was once dominant in Australian mineral sand processing, but improvements in spiral design — particularly the introduction of double-start spirals — have displaced it in most new installations, and few Reichert cone circuits remain in operation today.
Gold Sluice Boxes
The traditional sluice box used in alluvial gold mining consists of a long inclined trough fitted with transverse riffles or trapping mats that capture dense gold particles while lighter material is swept through. Modern recovery systems for artisanal gold mining often use high-recovery mats — carpet-lined, ribbed rubber or urethane matting — as the riffle medium, significantly improving gold recovery over traditional wood riffles. Sluice boxes scaled up for industrial use in placer mining operations can handle hundreds of tonnes per hour of gravel, making them among the highest-throughput gravity devices per unit of capital cost.
Mineral Applicability: When to Use Gravity Concentration
Gravity concentration is not universally applicable. Its effective deployment requires a sound assessment of the ore characteristics and a matching of those characteristics to the capabilities of available equipment. Several key factors determine whether gravity concentration is the appropriate choice and, if so, which equipment type is best suited.
Density Difference and Concentration Criterion
The concentration criterion (Δρ) provides the first-order filter for applicability. For Δρ above 2.5, gravity separation is straightforward and should always be the first option evaluated. For values between 1.75 and 2.5 — as encountered in some sulfide-gangue separations — gravity methods are viable but require good feed preparation and equipment selection. Below 1.75, gravity methods may be applied as roughing steps ahead of flotation or magnetic separation but are unlikely to produce finished concentrates. The absolute densities of the minerals involved also matter: separations involving high-density minerals such as gold, platinum, cassiterite, wolframite, or ilmenite generally give high concentration criteria, while separations of sulfide minerals (s.g. 4–5) from silicate gangue (s.g. 2.65–2.9) give moderate criteria that are workable with the right equipment and feed preparation.
Particle Size and Liberation
Gravity separation works best at coarser sizes where the mass-to-surface-area ratio of particles is high. For this reason, gravity circuits are ideally positioned as early as possible in the comminution sequence — as soon as adequate liberation of the target mineral is achieved, but before unnecessary fine grinding has reduced the particles to sizes where gravity methods lose effectiveness. Open-circuit rod milling, which produces a narrower and coarser size distribution than ball milling, is preferred over ball milling for feed preparation to gravity circuits when the target mineral is liberated at relatively coarse sizes. Where fine grinding is unavoidable, centrifugal concentrators extend the size range of gravity recovery.
Feed Variability and Slimes
Gravity separators are sensitive to variations in feed pulp density, and precise automatic density control is important for maintaining consistent performance. The presence of slimes — particles below about 10–20 μm — degrades separation performance by increasing slurry viscosity and physically coating mineral surfaces. De-sliming prior to gravity concentration is therefore standard practice, though it involves a loss of fine-particle values to the de-slimed fraction. Hydraulic classifiers are sometimes preferred over hydrocyclones for de-sliming gravity feeds because the lower shear forces generated by hydraulic classifiers cause less degradation of friable valuable minerals such as cassiterite or gold.
Sulfide Minerals
If the ore contains a significant proportion of sulfide minerals, these will tend to report to the gravity concentrate because of their relatively high density (pyrite 5.0, chalcopyrite 4.2, galena 7.5). If the primary grind is fine enough for sulfide flotation (below about 300 μm), it is standard practice to float the sulfides before gravity concentration of the remaining heavy minerals. If the grind is too coarse for flotation, the gravity concentrate must be reground and the sulfides removed by flotation before final upgrading of the non-sulfide heavy minerals.
Gravity Concentration for Gold Recovery
Gold is perhaps the highest-profile application of gravity concentration in the modern minerals industry. Its very high density (s.g. 19.3) gives a concentration criterion of over 11 against quartz gangue, making it the most amenable mineral to gravity concentration of any economically important metal. Yet despite this favorable characteristic, gravity recovery of gold has historically been underutilized at many operations where all gold extraction was assigned to cyanide leaching. A series of technological and economic developments has reversed this trend, and gravity recovery of free gold is now standard practice at most gold processing plants worldwide.
Gravity Recoverable Gold (GRG)
The key question in any gold gravity circuit design is: how much of the gold in the ore is amenable to recovery by gravity? The answer is quantified by the Gravity Recoverable Gold (GRG) test, standardized by Laplante et al. (1995). The GRG test subjects a sample to three successive stages of grinding and Knelson concentration, measuring the cumulative gold recovery and the gold distribution across particle-size fractions at each stage. The result is a GRG response curve that characterizes the proportion of total gold that is liberated as free metallic gold — as opposed to gold locked in sulfide minerals or in solid solution — at various grind sizes.
GRG values for gold ores range from nearly zero (for refractory ores where gold is entirely locked in arsenopyrite or other sulfides) to over 90% (for high-grade quartz-gold veins with coarse free gold). Typical placer deposits may have GRG values of 80–95%, while primary sulfide gold ores often have GRG values of 20–60%. The GRG result does not directly predict plant gravity recovery — which depends on circuit configuration, cyclone bypass, and other operational factors — but it is the essential input to simulation models used in circuit design. Plant gravity recovery typically ranges from 20–90% of the GRG test value.
In-Circuit Gravity Recovery
The most common configuration for gravity gold recovery in a modern processing plant places a Knelson or Falcon concentrator on the hydrocyclone underflow within the grinding circuit. The cyclone underflow is already enriched in dense minerals — including free gold — due to the density effect in hydrocyclone classification, and treating this stream allows recovery of free gold before it can be further comminuted to sizes where gravity methods lose efficiency. Because only a fraction of the cyclone underflow needs to be treated (the fraction that can be processed by the concentrator within each concentration cycle), the concentrator can be sized smaller than would be needed for a full-flow application, reducing capital cost.
The resulting gravity concentrate is typically processed by intensive cyanidation in a small intensive cyanide leach reactor rather than being directed to the main leach circuit. This allows the gold in the gravity concentrate to be extracted rapidly and efficiently in a small, secure, high-security environment, avoiding the security and accounting complications of handling a visible gold-bearing product through the main plant.
Flash Flotation Integration
In ores where a significant proportion of gold is associated with sulfide minerals — particularly pyrite — gravity concentration alone may leave considerable gold in the flotation feed. Flash flotation, in which a flotation cell is inserted directly into the grinding circuit to float liberated sulfide minerals from the mill discharge or cyclone underflow, is often used in combination with gravity concentration to maximize total gold recovery. The gravity concentrator captures free gold while the flash flotation cell captures gold-bearing sulfides; the two together can recover a large proportion of the total gold in the ore at very coarse particle sizes, before the material is exposed to the main flotation circuit. This combined approach has been shown to significantly improve overall plant gold recovery in operations with mixed free and sulfide-associated gold.
Artisanal and Small-Scale Mining
Gravity concentration remains the primary gold recovery method in artisanal and small-scale gold mining (ASGM), which accounts for approximately 20% of global gold production. The range of equipment used spans from traditional gold panning through basic sluice boxes to small-scale centrifugal concentrators such as the iCON (Falcon-derived) concentrator. A critical metallurgical and environmental issue in ASGM is the use of mercury amalgamation, which remains widespread because it captures fine gold that sluices miss. Improved gravity concentration equipment — specifically small centrifugal concentrators — offers a viable mercury-free alternative for most ASGM operations and is a focus of ongoing international development efforts.
Circuit Design and Integration with Other Processes
Effective gravity concentration circuit design goes well beyond selecting a piece of equipment: it requires an integrated view of the entire processing flowsheet, careful attention to feed preparation, and a strategy for managing middlings and interlocked particles.
Feed Preparation
Proper feed preparation is arguably more important to gravity circuit performance than equipment selection. The cardinal rules are: achieve adequate liberation before feeding to gravity separators; classify the feed into narrow size fractions; control feed pulp density precisely; and remove slimes. Liberation assessment by mineralogical examination or by metallurgical testing should establish the particle size at which target minerals are substantially liberated, and grinding should be targeted to this size — not finer. Over-grinding destroys the size advantage that gravity methods require and produces slimes that degrade performance.
Water balance management in the circuit is critical because gravity separators all have optimal operating pulp densities and are sensitive to deviations from these. Automatic density control on the raw feed, using nuclear gauges or other density sensors controlling water addition valves, is standard practice in well-designed circuits. Surge capacity and thickener or cyclone dewatering stations within the circuit smooth out variations in recycled water density caused by slimes build-up.
Multistage Circuits
For final concentrate production, multistage gravity circuits are almost universally used. A typical iron ore gravity circuit might use primary spirals for roughing, a secondary spiral circuit for cleaning the primary concentrate, and shaking tables for final upgrading of the cleaner concentrate. Middlings from each stage are either recycled to the appropriate upstream stage or directed to a regrind mill for liberation before retreatment. Each stage of regrinding should be designed to liberate the minimum additional mass required — additional grinding beyond the liberation size wastes energy, creates slimes, and may generate particles too fine for efficient gravity recovery.
The mathematical analysis of gravity circuits using partition curves provides a powerful design tool. As demonstrated for spiral circuits, the sharpness of separation of a rougher-cleaner combination with middlings recycle is always greater than for a single stage, and can approach the theoretical maximum separation efficiency when the number of stages and recycle loads are optimized. This analysis can be extended to complex multi-product circuits by matrix methods.
Integration with Flotation
Gravity concentration and froth flotation are complementary rather than competing processes, and many industrial circuits exploit both. Gravity concentration at coarse sizes captures liberated dense minerals early, reducing the load on the flotation circuit and improving overall recovery. Fine particles that gravity methods cannot treat efficiently are directed to flotation. Conversely, gravity concentration of flotation concentrates or flotation tailings can upgrade or scavenge materials that flotation has treated inefficiently. The hybrid approach of gravity plus flotation — treating the same stream in both unit operations with the products combined — is increasingly used for gold recovery in complex ores.
Environmental Considerations
Gravity concentration circuits, when properly designed and operated, generate simpler and more environmentally benign waste streams than flotation circuits. Tailings from gravity operations are generally free of flotation reagents and, for dry tailings disposal or paste tailings applications, are often easier to handle than flotation tailings. Water recycling within gravity circuits is straightforward because the water contains no dissolved reagents that would interfere with reagent dosing in downstream processes. The lower energy consumption of gravity concentrators relative to flotation also contributes to a smaller environmental footprint per tonne of concentrate produced.
References and Further Reading
- Wills, B.A. & Finch, J.A. (2016). Wills’ Mineral Processing Technology, 8th Edition. Butterworth-Heinemann/Elsevier.
- Burt, R.O. (1985). Gravity Concentration Technology. Developments in Mineral Processing Series, Vol. 5. Elsevier, Amsterdam.
- Laplante, A.R., Woodcock, F. & Noaparast, M. (1995). Predicting gravity separation gold recoveries. Minerals and Metallurgical Processing, 12(2), 74–79.
- Bam, L.C., et al. (2023). Physical beneficiation of heavy minerals — Part 1: A state of the art literature review on gravity concentration techniques. Heliyon, 9(8). PMC.
- Edraki, M., et al. (2024). Gravity Concentration of Gold-Bearing Ores and Processing of Concentrates: A Review. Mineral Processing and Extractive Metallurgy Review. Taylor & Francis.
- Ancia, R., et al. (2020). Research and application of a Knelson concentrator: A review. Minerals Engineering. ScienceDirect.
- Sepro Mineral Systems: Gravity Concentration — Equipment Selection Guide.
- Multotec: How Do Spiral Concentrators Work?
- 911Metallurgist: Gravity Spiral Concentrator Working Principle.
- ScienceDirect Topics: Spiral Concentrators — Overview.