Gold Ore Crushing: Fine Is Not the Same as Liberated
When a gold plant underperforms, the crushing circuit is often asked to make the product finer. That is usually the wrong lever. Crushing in a gold flowsheet has one job — deliver particles the next stage can treat economically — and past a certain point, extra size reduction produces slimes that cost you recovery, reagent and throughput without liberating anything more. The circuit is a feeding device, and it should be judged on what the mill receives, not on how small it can go.
This article covers where over-crushing is paid for, how the last crushing stage should be chosen, why ore variability rather than the average grade is the design case, and what to check before you accept a flowsheet.
I. Liberation, Not Fineness
Gold is not evenly distributed through ore. It sits in veins, on grain boundaries, locked in sulphides or adsorbed on clays. The size at which the valuable mineral is physically exposed is a property of the deposit, and it does not change because the crusher is set tighter.
• Liberation is a measured property. It comes from a mineralogical examination of the ore, not from a crusher setting. Without that examination, "crush finer for better recovery" is an assumption, not an engineering decision.
• Size reduction past liberation adds cost and removes nothing else. The energy goes in, the slimes are created, the downstream stage has to handle them, and the recovery curve does not move.
• And the downstream stage has its own preferred feed. A heap leach needs permeability; a CIL or CIP circuit needs particles that stay in suspension and leach at a workable rate; gravity recovery wants the coarse free gold out early, before it can be slimed. Those three want different crush sizes, and none of them is "as fine as possible".
• So the circuit brief is a compromise, written down. Ask for the target size, the tolerance around it, and the reason — then design the crusher stages to hit it repeatably, including when the ore changes.
II. Where Over-Crushing Is Paid For
• Slimes are a recovery problem before they are a handling problem. Very fine particles leach slowly and are harder to separate from solution. In a heap they migrate, block the permeability of the bed and channel the solution away from the ore it was meant to contact.
• Reagent consumption is a function of surface area. More fines means more surface for cyanide and oxygen to be consumed by gangue minerals rather than by gold. The cost shows up in the reagent account, not in the crusher's power bill.
• Thickening and filtration get harder. A feed with more slimes produces a slower-settling slurry, which means more flocculant, bigger thickeners or a lower throughput — a capital consequence of a crushing decision.
• Free coarse gold can be lost in the circuit itself. If gravity recovery is part of the flowsheet, crushing to a fine product before the gravity unit means the coarse gold is slimed and the gravity unit never sees it.
• And energy per tonne rises faster than size falls. The last increment of reduction is always the most expensive, which is why the crush-grind split matters more than the total ratio.
III. Choosing the Last Stage
| Option | Fines generated | Suits | Watch |
| Jaw + cone (two or three stage) | Low to moderate | Most hard rock ores; predictable product | Cavity choice and closed-side setting control the result |
| Jaw + cone + VSI | High | Where shape matters or the ore is friable | Slimes in the leach feed; checks the downstream stage can cope |
| Jaw + HPGR | Low, with micro-cracking | Hard ores where downstream grinding is the bottleneck | Capital, wear surface cost, and feed preparation demands |
| Jaw + hammer or impact | High and variable | Soft, low-abrasion ore only | Wear cost per tonne and a wider product spread |
The table is not a ranking. The cheapest circuit to buy is rarely the cheapest to run, and the option that makes the finest product is usually the one that creates the most slimes. The right question is which stage the plant is short of — if the mill is the constraint, crushing finer may pay; if the leach or the thickener is the constraint, it will not.
• Established, competent, abrasive ore usually belongs in a cone at the end, because it gives a controlled product with the least fines.
• Friable or clay-rich ore behaves badly in a compression chamber; the choice is about handling, not about reduction.
• And a circuit that cannot hold its setting — because the liners wear, or because the feed varies — will not deliver the size the flowsheet assumes, whatever the stage.
IV. Design for the Ore You Will Actually Mill
A single composite sample describes a deposit that does not exist. Every gold circuit lives with ore that changes — harder, softer, more clay, more sulphide, different gold association — and the crushing circuit is where that variability first becomes a problem.
• Hardness variability changes the product size. The same setting on a softer ore gives a finer product and more fines; on a harder ore it gives coarser material and a lower throughput. Neither is what the flowsheet assumed.
• Clay and moisture change the flow. Sticky ore bridges in hoppers, blinds screens and slows the circuit. A plant designed on dry, clean samples will spend its first wet season at half rate.
• The design case is the bad week, not the average month. Ask what the circuit does when the ore is at its hardest, wettest and most clay-rich, and whether the answer is a rate reduction or a stoppage.
• Sampling is the foundation. A flowsheet built on a grab sample is a guess with a price list attached. Insist on sampling that represents the orebody, and testwork at the size range you will operate in.
V. Where the Split Between Crushing and Grinding Belongs
Crushing is cheaper per tonne of reduction than grinding, which is why the industry pushes work upstream. That rule has a limit: the point where extra reduction starts making fines the next stage cannot use.
• Move reduction into crushing only while it is still making usable particles. A crusher that converts a third of its output into a fraction the leach cannot handle has not saved the mill any work; it has moved the loss upstream.
• Compare on the total circuit, not on kWh per stage. The saving at the crusher is real; so is the cost at the thickener, the leach and the reagent account — and it lands in a different budget line, which is why the two are never compared.
• Ask which stage the plant is short of. If the mill is the constraint, more crushing pays. If the leach, the thickener or the filtration is the constraint, more crushing makes the bottleneck worse while looking like an improvement.
• And leave room to change your mind. Ore changes over a mine's life, and the split that was right at start-up is often wrong at depth. A circuit with room to insert a third stage beats one that must be rebuilt.
VI. What to Ask Before You Accept the Flowsheet
• What is the target size at the crusher discharge, and what is it based on? A mineralogical liberation study, the leach testwork, or a rule of thumb — the answer tells you how much confidence to place in the number.
• What tolerance does the downstream stage accept? Coarse and fine limits both, because either extreme causes a problem and the circuit has to stay inside the window.
• How is the setting held, and how is it measured? Closing-side setting, liner wear allowance and the inspection interval together decide whether the design size survives the first month.
• What happens at the hardest and wettest ore? Ask for the rate and the product size in both cases, not just at the design point.
• Where does the recirculating load go, and what limits it? A circuit without a clear answer here has an unplanned bottleneck in its future.
• What testwork supports the fines estimate? If nobody has measured how much material will be produced below the smallest size the next stage handles, the slimes problem has not been designed for.
Frequently Asked Questions About Gold Ore Crushing
Q: Will crushing finer improve my recovery?
A: Only if the gold is still locked at your current size. Once liberation is reached, finer crushing adds slimes, reagent consumption and downstream load without releasing more gold. The evidence you need is a liberation study, not a trial run.
Q: Is a two-stage circuit enough?
A: It depends on the reduction the ore allows and the size the next stage wants. Two stages are common where the mill does the fine work; a third stage earns its cost when the mill is the bottleneck or the leach demands a tighter feed.
Q: Where should gravity recovery sit?
A: Early, before the gold can be slimed — usually in the grinding circuit rather than after fine crushing. If coarse free gold is present, the crushing circuit should not be making it finer than the gravity unit can catch.
Q: How important is the crusher's product shape?
A: Less than in aggregate production, more than most flowsheets assume. Flaky particles pack and reduce heap permeability, and they behave differently in a mill. Shape is second-order in gold, but not zero.
For crushing equipment configured around a gold plant's target size and ore variability, see the crushing equipment range.
