Best Gold Mining Equipment
From Dirt to Gold Bars

Free gold and refractory gold need completely different equipment routes. Free gold hits high recovery through gravity separation. Refractory gold may need oxidative pretreatment and cyanide leaching. Get this call wrong and everything else is a write-off. Small operators have dropped serious money on full CIL circuits when their gold was coarse and free, a jig could have handled it. The reverse happens too, people running tables and sluices into the ground while recovery sits at 65% because the gold is sulfide-locked.

Polished section work sometimes shows free gold while gravity recovery stays flat. One cause: surface condition. Gold that has been sitting in the oxidation zone accumulates a film of iron-manganese oxide that narrows the density differential against gangue. A scrubbing step before gravity separation strips the film. No equipment swap needed.

Another cause: particle shape. Dendritic gold settles far slower than spherical gold of the same mass. Behaves more like a light mineral in a gravity field. Centrifugal concentrators handle it much better than tables or jigs because the enhanced-g force overrides the shape factor. Gold morphology analysis gets skipped in met test programs constantly.

Throughput. Five tons a day versus five thousand, completely different equipment logic. Vendors push one size larger, always, citing expansion headroom. A ball mill oversized by 30% drags along bigger foundations, heavier cabling, a bigger transformer, a higher ball charge, and fatter power bills. The cascade cost can run one and a half to two times the equipment price. Select to confirmed reserves, not to the geologist's blue-sky number.

Water, power, altitude. These override equipment specs more often than people expect. At altitude, diesel engines lose grunt, flotation aeration drops, bearings run hotter. Water quality matters more than water quantity. Thiosulfate in groundwater chews through cyanide. Organic-rich surface water fills flotation cells with barren froth. Acidic water corrodes everything. Hard water scales pipes and clogs air lines. A full water quality analysis before equipment selection, not after. The lab bill is trivial.

Forty to sixty percent of plant energy goes into making rocks smaller. Operating cost lives here.

Jaw crushers. Almost nothing breaks on these. Discharge gap adjustment range matters more than nameplate capacity. Jaw plate material matters more than model. High-manganese steel plates on quartz vein ore wear out shockingly fast. High-chrome cast iron plates go two to three times longer.

A suspended permanent magnet over the feed opening catches tramp iron. Stray drill bits, blast pipe bits, scraps of old rail. One chunk in the jaw cavity cracks a plate or snaps the eccentric shaft. Most equipment lists do not include this.

Cone crushers must be choke-fed. Running one starved causes uneven liner wear. Cavity type (standard, medium, short-head) is critical for product size distribution; vendor literature covers this adequately.

Steel ball quality. Low-chrome balls break at a high rate. The broken fragments jam grate plates and consume energy doing nothing. High-chrome balls (north of 10% Cr) wear at about 60 to 70% the rate of low-chrome and break far less often. Higher unit price, frequently lower total cost. Ask for drop-ball impact test data from working mines, not just the catalog.

A system issue that gets blamed on the mill all the time: classification efficiency. The ball mill and its hydrocyclone form a closed loop. Poor classification sends already-ground material back for regrinding. Waste of energy, plus overground particles are terrible in flotation: high surface area, reagent-hungry, poor bubble attachment. Operators tweak the ball charge, swap liners, adjust speed, recovery does not move. Swap in a properly matched cyclone (spigot and vortex finder diameters) and both grinding efficiency and downstream recovery improve at once.

HPGR. Two counter-rotating rolls, extreme pressure, dense micro-fracture networks in the product. Unit energy 20 to 30% below ball milling. Sensitive to feed moisture above 3%. The micro-fractured flake lets the downstream ball mill hit target recovery at a coarser grind, which means lower ball mill load and energy draw. This indirect benefit almost never shows up in equipment comparison spreadsheets, which is a shame because it can be substantial.

More space here than elsewhere. Gravity separation is where the information gap between what operators need and what they have access to is widest. Flotation and cyanide leaching at scale have been standardized by major miners with engineering consultancies backing them up. Gravity circuits for small and medium operations remain largely trial-and-error. Equipment selection mistakes cluster here disproportionately.

One gravity device for a small mine: a jig.

Tables have better precision. Centrifuges catch finer gold. Spirals use no power. The jig wins none of these categories and wins on the one that matters most at small scale: forgiveness. Feed concentration swings 10%, jig keeps producing. Same swing on a table and the banding scrambles. Small mines cannot hold conditions steady. Water pressure wanders, the excavator alternates between coarse and fine scoops, the night shift is greener than the day shift. A machine that performs acceptably under rough conditions beats a machine that performs beautifully under perfect conditions.

Sawtooth jig: asymmetric pulse, fast rise, slow fall. Long stroke low frequency for coarse, short stroke high frequency for fines. Standard knowledge.

What gets less coverage: the same jig processing the same ore can show different results day shift versus night shift, and operator skill is not the reason. Water supply pressure. Lower usage at night means higher line pressure, which changes the pulsation waveform. A constant-pressure header tank on the supply line fixes it. Costs very little.

Bed management on sulfide gold ore: hematite particles as artificial ragging, density between gangue and gold, creating a graded buffer. Ragging grain size slightly coarser than the feed. Too fine, it washes out. Too coarse, gangue bleeds through the gaps. Impact on recovery can be five to ten percentage points.

Cold starts. Every time the jig starts up, there is no bed. Gold passes through until one forms, half an hour to an hour. Pre-loading coarse lead particles (2 to 3 mm) on the bottom as a false concentrate layer lets effective separation begin immediately. Small mines that start and stop multiple times daily hemorrhage gold during these bed-establishment windows without realizing it.

Manual versus automatic discharge. Manual discharge valves require the operator to judge timing by the look and feel of the concentrate stream. Takes three to six months to develop this eye. Gold lost during the learning period. Automatic valves running off bed pressure sensors add 10 to 15% to the price of a medium jig. Where operator turnover is high, they pay for themselves.

Cleaning duty only. A 6-S table processes a ton or two per hour. Roughing goes to jigs or spirals; table handles the rougher concentrate.

Cross-slope angle and wash water flow interplay is the core skill. Experienced operators read the table by sound. Dull heavy rumble: bed too thick, more wash water. Thin bright rattle: material running lean, increase feed density. After an adjustment, fifteen to twenty minutes before touching anything again. The table needs that long to settle into a new equilibrium. The overwhelming temptation is to keep tweaking the water valve while watching the gold band. That makes things worse, every single time.

Installation detail that gets missed: the deck must be level longitudinally, dead level. A fraction of a degree sends material off its design flow path. Small-mine tables often sit on platforms that were not level when they were built and get worse as vibration loosens bolts. Weekly check with a spirit level.

Knelson: fluidized bed, water injection controls concentrate bed thickness, parameter-sensitive, high precision. Falcon: smooth bowl, no fluidization water, simpler, higher throughput, lower concentrate grade. Knelson for higher-grade narrow-size ores, Falcon for low-grade volume work.

Install in the mill circuit circulating load, where gold concentration runs three to five times fresh feed.

Knelson fluidization water should track feed heavy mineral content, which fluctuates constantly while water pressure is usually set and forgotten. Linking the water valve to a slurry density meter is a straightforward retrofit that keeps the two in step.

Flushing interval. Too long and the concentrate bed fills, gold starts going to tails. Too short and throughput suffers plus every flush has a bed re-establishment dead zone. No universal number. Run the first few days at varied intervals (30, 60, 90, 120 minutes), weigh and assay each flush separately, plot gold per unit time against interval. Peak of the curve is the answer. Two to three days of testing, permanent gain.

Zero power, zero moving parts, zero maintenance. Lower precision than tables or centrifuges. Roughing and scavenging. Where electricity is expensive, spirals plus jigs is often the cheapest circuit that still functions.

Feed density swings hit spiral performance hard. Too thick and material crawls down the trough, heavies do not reach the inner edge before hitting the tails port. Too thin and slurry runs in a narrow center ribbon, wasting most of the separation surface. Somewhere around 25 to 35% solids by mass. A small buffer box with an overflow weir on the feed line, keeping density roughly steady, does more than chasing spiral brand names.

Three parts equipment, seven parts reagent regime and operator control. Same cell, different collector, different pH, 20-point swing in recovery. So less about hardware here, more about what surrounds it.

Mechanical cells. Air volume control and bubble size uniformity outweigh cell volume. Rotor-stator gap affects mineralization. Too wide, energy dissipates before the bubble picks up the particle. Too narrow, shear rips particles back off.

Flotation columns. Better grade uplift in cleaning than conventional cells. Not much else to add that is not already in the vendor specs.

pH modifier. Lime is cheap and universal. Too much lime depresses pyrite. Also forms gypsum scale inside cells and on air lines, slowly strangling aeration. Soda ash is a few cents more per kilogram. Overall performance is generally better. On arsenical gold ore specifically, soda ash interferes less with arsenopyrite floatability. This matters because a lot of gold rides with arsenopyrite, and depressing it means depressing gold recovery.

Water temperature. Below about 5 degrees C, collector adsorption kinetics slow, frother gets sluggish. Same reagent recipe works in summer, goes flat in winter, and nobody changes anything so everyone assumes the ore changed. Ball mill discharge runs warm from grinding friction. A heat exchanger transferring that warmth to flotation feed water solves the seasonal swing at essentially zero added energy cost.

CIL and CIP. CIL runs leaching and adsorption simultaneously. CIP leaches first, adsorbs second. Preg-robbing ore (natural carbonaceous matter) must go CIL. Non-carbonaceous ore is more economical on CIP.

Coconut shell carbon, not coal-based. In CIL, coal carbon breaks down fast. Gold exits on fine carbon fragments in the tailings. Standard tailings assay does not tell you whether gold is on mineral surfaces or on carbon fines. The loss is invisible. Plants with chronically elevated tailings grade that nobody can explain: fine carbon is worth investigating as a cause.

Carbon screening frequency. Carbon moves counter-current through the tank train, screened at each transfer to remove fines. Plants frequently stretch the interval from daily to every second or third day to save downtime. Fine carbon sitting in the circuit longer desorbs gold back into solution. Fine carbon desorbs faster than intact granules because of the surface-to-volume ratio and abrasion exposure.

Tank count. Six to eight is the textbook number. Carbon travels counter-current: freshest carbon in the last tank contacts the weakest gold solution, which maximizes the driving force in each stage. Fewer tanks, the discharge still carries too much gold. More tanks, the later stages do almost nothing while consuming floor space and motor power. The number needs to come out of leach kinetics data and carbon isotherm data for the specific ore. "Eight tanks because that is what everyone else does" is not engineering, but it shows up in feasibility studies.

Heap leaching. Low-grade ore. Crushed, stacked, irrigated with cyanide from the top, pregnant solution collected at the base.

Agglomeration drum. Ore with significant fines must be agglomerated before stacking. Without it, fines plug percolation channels. Solution distribution goes patchy, channeling and ponding develop.

In-heap precipitation. Gold dissolves, starts migrating downward in solution. If pH drops locally (acidic minerals) or dissolved oxygen gets consumed en route, gold drops back out of solution and plates onto ore surfaces inside the heap. Gone permanently. Cannot be re-leached. This is why heap height matters. Above roughly 10 meters, solution reaching the base has lost enough chemical potency that gold dissolution is weak and gold that dissolved higher up re-precipitates on the way down. Stacking to 15 or 20 meters saves land and sacrifices the bottom zone.

Irrigation system. Wobbler sprinklers are cheap and common and heavily affected by wind. On a windy day, irrigation intensity across the heap surface can vary by a factor of two or three. Some zones overwashed, adjacent zones nearly dry. Drip irrigation, PE tubing in a dense grid on the surface, distributes solution far more evenly. In arid windy areas drip also loses much less to evaporation. Where water is scarce that benefit alone can justify the switch.

Pregnant solution goes to zinc precipitation (Merrill-Crowe) or carbon adsorption. Merrill-Crowe needs the solution deaerated and clarified to a standard that is much tighter than what CIL or CIP requires. Suspended solids or dissolved oxygen in the feed and zinc consumption spikes, precipitation efficiency drops. The deaeration tower and clarifier for a Merrill-Crowe circuit cannot be cheaped out.

Placer gold is free gold, already liberated. No crushing, no grinding. The work is separating it from sand and gravel.

Dredges. Digging, washing, gravity separation, all on one floating platform. Bucket line dredges move thousands of cubic meters a day. High capital, strict requirements on water depth, channel width, bedrock profile.

Bucket chain speed should stay below about 0.35 meters per second. Above that the buckets create enough turbulence to scatter flat gold off the bedrock surface and carry it downstream. Unrecoverable. This loss never shows in any production report because it was never captured, therefore never measured. Engineers who have set up sampling stations downstream of dredges to measure background gold concentration have found the numbers uncomfortable.

Mobile wash plants and trommels. Feed hopper, trommel screen, sluice or jig, all on a trailer. Placer deposits are patchy, frequent moves are a given. Screen aperture needs to match gold particle size distribution. Most of the gold under one millimeter and the aperture too coarse: gold-bearing fines leave with the oversize.

Sluice lining. Riffled rubber mats and carpet handle coarse gold. They are poor on fine gold and flake gold. Hungarian riffles and miners moss are materially better for fines. A sluice lining is not an expensive item. The sluice is the last separation step. Everything upstream feeds into it. Wrong lining and the upstream effort is wasted.

Sluice cleanout frequency. Depends on heavy mineral content in the feed. High heavies: riffle spaces fill up in two to three hours, new gold arriving starts washing over the top. Low heavies: six to eight hours is workable. A sampling tray at the tail end, checked periodically for black sand content, shows when retention is declining.

Dig face management. Not equipment per se, but it determines how much gold reaches the equipment. Gold in alluvial deposits concentrates in the bottom few tens of centimeters above bedrock. If the excavator stops short of bedrock, the richest layer stays in the ground. If bedrock has cracks and hollows, gold accumulates in them. Some operations wash exposed bedrock with a high-pressure jet after excavation to flush crevice gold into the wash plant. On irregular bedrock the production bump from this step is significant.

Gravel pumps. Gravel is savagely abrasive. Standard pumps get eaten through in days. High-chrome or rubber-lined only. Keep three months of impellers and wear plates on hand. A gravel pump down means the whole operation stops, and shipping a replacement part takes weeks.

Handheld XRF. Rapid field grade check. Gold detection limit around 2 to 5 ppm. A lot of mineable gold deposits run below 1 ppm. Workaround: measure arsenic and antimony instead. In many gold deposits As and Sb correlate tightly with Au. Measuring pathfinder elements to infer gold enrichment extends the XRF's effective reach.

Heavy liquid separation. Tribromomethane or diiodomethane, density fractionation in the lab. Shows gold liberation, distribution across density classes, gravity recoverable gold index. Do this before buying any gravity equipment. A single heavy liquid test contains more useful selection data than a stack of equipment catalogs.

Metal detector. Pulse induction for gold nugget detection. The critical selection factor is how the unit handles ground mineralization. Magnetite-rich or hydrothermally altered ground generates a wall of false signal. VLF units are useless in that environment. PI units with continuous automatic ground balance tracking are necessary. Always test on the actual ground, not in a showroom.

Slurry pumps. Operating point belongs in the left third of the efficiency curve, not at the peak. Slurry properties drift. Concentration rises, resistance increases, operating point shifts right. Starting at peak means any shift puts the pump into the unstable zone.

Thickener. Flocculant injection point: in the feed pipe, using pipe turbulence for mixing. Not into the tank body. Floc structure and settling speed are both better with pipe injection.

Filter press. Tailings below 20% moisture, dry-stackable. In a growing number of jurisdictions, no dry stack means no environmental permit.

At any given moment, a running processing plant has gold sitting in ball mill liner grooves, classifier bottoms, flotation false bottoms, pump casing dead spots, pipe elbows, thickener rake bases, concrete floor cracks. Does not appear in daily reports. Surfaces during shutdowns and cleanouts. A 10,000-ounce-per-year plant typically holds somewhere between two hundred and five hundred ounces inside its equipment and infrastructure at any given time.

Flushing mills, classifiers, and pipes thoroughly during scheduled shutdowns recovers gold that often covers the maintenance bill.

Old plant floors. Decades of slurry splashing soaks gold-bearing material into cracks and concrete pores. Plants that ran for ten-plus years, when demolished, yield rubble assaying at tens to over a hundred grams per ton. Some decommissioned operations have recovered worthwhile gold just from ripping up and processing the floor slab. Implication for new construction: seamless epoxy floor coating, sloped to drain all spills into collection channels.

Pipe layout. Minimize elbows. Every elbow accumulates gold. Heavy particles in slurry hit the outer wall at bends, decelerate, settle. Straight runs accumulate much less. Where bends are necessary, large-radius elbows over standard short-radius. Some plants fit removable inspection ports at critical elbows for periodic cleanout.

Circuit recovery follows its weakest link. Flowsheet first, mass balance, bench tests.

Equipment hits stride three to six months after commissioning, not at startup. Every machine in its particular ore has a break-in phase. Pushing for design targets in the first weeks triggers over-correction and instability.

Sampling gear gets left off every purchase list and matters more than any single processing unit. An on-stream analyzer at key circuit points, providing continuous grade data, typically pays itself back in a year or two through optimization.