P80 was first introduced on how gold tailings are formed as the single grinding parameter that most determines liberation. This page is where that concept gets a full, dedicated treatment as a testwork question in its own right, since almost every recovery method and equipment choice covered elsewhere in this set ultimately depends on getting this one variable right.
What does "liberation" actually mean?
Liberation means a gold particle has been physically freed from the surrounding gangue or sulphide it was originally locked inside, so that a recovery process, gravity, flotation, or leaching, can actually act on the gold itself rather than on a composite particle that happens to contain some gold along with a lot of worthless rock. A particle can carry a very high gold grade and still be economically worthless if it remains unliberated, because no recovery method can reach gold it can't physically get at.
Liberation is not a binary, all-or-nothing state either. A particle can be fully liberated, with gold entirely exposed at its surface; partially liberated, with some gold exposed and some still enclosed; or fully locked, with no gold surface exposed at all. Quantifying that distribution across a whole sample, not just estimating an average liberation percentage, is what a proper mineralogical liberation study actually provides, since a dump with a small population of fully locked coarse particles and a large population of fully liberated fine particles behaves very differently in a recovery circuit than one with everything sitting at some intermediate, partially liberated state.
How is liberation actually measured in the laboratory?
Automated mineralogical analysis, most commonly QEMSCAN or equivalent scanning electron microscope-based systems, is the standard modern method for quantifying liberation directly, since it images polished particle sections and classifies the minerals present at a resolution fine enough to see exactly how much gold surface is exposed versus enclosed. This produces a genuine liberation-by-size-fraction dataset rather than an inference from grade alone, which is why liberation studies are treated as a distinct testwork line item from a standard fire assay.
Older or lower-budget liberation assessments sometimes rely on optical microscopy of polished sections combined with manual point counting, which can still produce a usable liberation estimate but with more analyst time and generally coarser resolution than an automated scan provides. Either method is a real improvement over assuming liberation from grade or particle size alone, but the automated route has become the industry default where budget allows because of the speed and statistical robustness it offers across a large number of particles.
How is P80 measured and controlled?
P80 is measured through sieve analysis for coarser material or laser diffraction for finer fractions, and a plant's grinding circuit is tuned to hit a target P80 by controlling mill residence time, ball or rod charge, and classifier cut point. The target itself is not arbitrary: it comes out of liberation testwork on the specific material, since grinding finer than the liberation size a dump actually needs simply wastes energy without recovering more gold.
Overgrinding is a genuine cost, not merely a missed efficiency. Grinding energy consumption rises sharply as target particle size decreases, so a plant grinding well past the point where liberation is actually achieved is spending real money and generating real wear on grinding media for no additional recovery benefit, and in some cases can even harm recovery by generating excess ultrafine slimes that are themselves harder to recover by gravity or that increase reagent consumption downstream. Hitting the correct P80, not simply grinding as fine as practically possible, is the actual optimization target.
How does particle size interact with recovery method choice?
Coarse, liberated gold suits gravity concentration directly. Fine, liberated gold needs leaching, since it's too small for gravity equipment to catch efficiently, a problem covered in full on fine gold recovery: the sub-75-micron problem. Unliberated, locked gold needs finer grinding or a liberation-improving step regardless of the grade reported, because no recovery method downstream can compensate for gold that grinding never freed in the first place. Confirming which of these three situations a dump is actually in is exactly what the GRG test is built to answer for the gravity-recoverable fraction specifically.
A single tailings dump commonly contains all three populations simultaneously in different proportions, which is why a flowsheet design usually combines gravity, leaching, and sometimes a regrind stage rather than betting everything on one recovery method. The relative proportion of each population, established by the same particle size and liberation testwork this page describes, is what actually determines how much capacity a project should build into each stage rather than a generic industry rule of thumb.
How does a regrind stage actually improve liberation?
A regrind stage, covered in full on regrind milling: when tailings need finer grinding, takes a specific stream, usually a rougher concentrate or a coarse middlings fraction rather than the entire plant feed, and grinds it to a finer target size specifically to break open composite particles that were only partially liberated at the original grind. Regrinding the whole feed stream rather than a targeted, already-upgraded stream would be far more energy-intensive for the same liberation gain, which is why regrind circuits are almost always sized for a reduced-volume stream downstream of an earlier concentration step rather than the full tonnage.
Ball mills and, increasingly, stirred media mills are the common equipment choices for regrinding, with stirred mills generally favored for very fine regrind targets because they achieve fine grinding more energy-efficiently than a conventional ball mill at that size range. Selecting between them is, again, a testwork and economics question specific to the target grind size and the tonnage involved, not a default equipment choice.
Why does this matter more for tailings than fresh ore?
Tailings arrive at a grind size that was already fixed by the original plant, often decades ago, and that grind can't be changed without regrinding the material, which is itself a real capital and energy cost rather than a design choice made freely at the outset. A fresh-ore project chooses its target P80 from a blank sheet; a tailings retreatment project inherits whatever grind the historic plant left behind and has to decide whether paying to regrind is worth the additional liberated gold it would recover, a plant-design decision covered on gold processing plants for tailings.
The historic grind a legacy Ghanaian dump was left at also varies considerably depending on the era and scale of the original operation, since grinding technology and target liberation standards have themselves evolved over the decades these dumps span. A tailings retreatment project therefore cannot assume a specific historic P80 without measuring it directly on the dump in question, since two dumps of similar age and origin can still have been ground to meaningfully different targets by their respective original operators.
When does regrinding old tailings actually pay for itself?
Regrinding pays for itself when the additional locked gold it liberates is worth more than the extra energy and equipment cost of finer grinding, an economic comparison that depends entirely on how much gold liberation testwork shows is actually locked at the current grind versus a finer one. It is never assumed as a default step; it is a decision made from real data, the same testwork-first principle running through this entire recovery methods set.
How this connects to deportment studies
Particle size and liberation testwork runs alongside, not instead of, the deportment work already covered on gold tailings composition, since knowing a gold form (free, locked, surface-bound) is only half the picture without also knowing the particle size it occurs at.
Why this testwork belongs early in a project timeline, not late
Because particle size distribution and liberation results feed directly into equipment sizing decisions across gravity, regrind, and leach circuits alike, this testwork needs to happen early enough in a project's technical program to actually inform flowsheet design, rather than being run late as a confirmation exercise after major equipment decisions are already locked in. A project that reverses this order risks specifying equipment first and discovering only afterward that the feed's actual liberation profile called for a different configuration.
What this means for investors reviewing a project's technical case
A feasibility study that states a target recovery figure without showing the underlying particle size distribution and liberation data behind it is presenting a conclusion without its supporting evidence. Investors evaluating a Ghanaian tailings retreatment opportunity are entitled to ask for this data specifically, since it is the direct technical basis for every equipment and flowsheet decision the rest of the project's economics depend on.