The grind–leach handshake: how size, HPGR, and chemistry lift gold and copper recovery

On paper, it looks straightforward: grind to the liberation size, leach, and recover. In practice, the interface between the mill and the leach tank determines whether ounces depart in doré or in tailings. Full‑scale and pilot data show how the right comminution choices — including High‑Pressure Grinding Rolls (HPGR, a press that induces micro‑cracks and high fines) — and a disciplined leach window shift both kinetics and economics.

Case in point: feeding HPGR product to ball mills has increased mill throughput by about 27–44% (from roughly 203 to 254–295 t/h) at constant power while dropping Bond Work Index from 11 kWh/t to about 8–9 kWh/t (scielo.org.za). For copper leaching, similar HPGR test programs recorded a 2–10% bump in Cu extraction versus conventional crushing (ui.adsabs.harvard.edu), echoing industrial experience at Los Colorados (Chile), where HPGR fines enabled about 30% higher ball‑mill capacity at lower energy (scielo.org.za).

At the other end of the flowsheet, leach performance tracks particle size: in gold cyanidation, ≥90% of gold in the finest lab fraction (<20–37 μm) dissolved without oxidation pretreatment, while coarser classes lagged (mdpi.com) (mdpi.com). And thickening the tails quickly, rather than allowing cyanide in slimes ponds to volatilize or oxidize over days, can make cyanide loss negligible compared with the ~90% loss observed in unthickened ponds (911metallurgist.com).

Grinding circuit selection and size reduction

Conventional gold circuits — SAG/AG plus ball mills — commonly target P80 (the particle size at which 80% of the material is finer) of about 75–150 μm to liberate gold; ultrafine mills (IsaMill/Vertimill) can push below 20–50 μm for refractory ores (mdpi.com) (mdpi.com). Copper heap leach ores typically stay much coarser (top size around 10–50 mm) to preserve permeability, while copper flotation concentrates are often ground to roughly 20–75 μm before pressure leach.

HPGR as a pre‑crusher shifts the downstream balance: finer products and micro‑cracks cut mill energy and raise recovery. Full‑scale trials recorded ~27–44% higher ball‑mill throughput at equal power with Bond Work Index dropping from 11 to ~8–9 kWh/t (scielo.org.za), with industrial echoes at Los Colorados (about 30% higher capacity) (scielo.org.za).

Classifier control matters: in closed circuits, cyclone or screen cut sizes govern what enters leach. Ensuring only sufficiently fine overflow goes forward minimizes locked metal. For example, controlling ball‑mill cyclone overflow at 80% <100 μm helps target gold liberation; minimizing oversize recycle avoids energy spent with little added leach gain once the liberation threshold is met (mdpi.com) (scielo.org.za). In practice, operating to just‑sufficient fineness (plus targeted HPGR) delivered >90% gold recovery for the <0.02–0.037 mm fraction (mdpi.com), and further grinding of that fraction did not significantly increase leach recovery (mdpi.com).

Particle size effects on leaching kinetics

Gold cyanidation shows strong size dependence: dissolution rises markedly as particle size decreases, with ≥90% of gold dissolving in the finest lab fraction (<20–37 μm) without oxidation pretreatment (mdpi.com) (mdpi.com). Once target size classes are achieved, longer mill time for that size class does not deliver significant extra leach recovery (mdpi.com).

For copper, coarser grinds are common: heap leach ore at about 10–50 mm and concentrates for autoclave/pressure leach at roughly 25–45 μm P80. Even so, fines and micro‑fractures from HPGR improve bacterial and acid leach rates; replacing tertiary crushing with HPGR raised sulfide copper (SCC) leach extraction by 2–10% in bench and pilot tests (ui.adsabs.harvard.edu).

Slurry properties and thickening for leaching

Pulp density (solids concentration) is a primary lever. Modeling shows that increasing solids from, for example, 40% to 50% w/w lowers flowrate and increases retention time in each leach tank (scielo.org.za). Many gold CIP/CIL (carbon‑in‑pulp / carbon‑in‑leach) plants operate at 45–50% solids to maximize throughput, consistent with such modeling; the optimum is the highest density that still allows effective mixing and oxygen transfer.

Speed matters in tailings handling. In unthickened slimes ponds, residual cyanide losses of ~90% over days are reported, driven by volatilization and oxidation, whereas high‑rate thickeners (HRT) with about 3 h retention make cyanide loss negligible by recovering water quickly (911metallurgist.com). Plants commonly reach ≥40–50% solids underflow with well‑chosen polymer dose; high‑molecular‑weight flocculants form large aggregates, enabling “much higher mass throughput” through a given thickener area (research.csiro.au). In practice, with tuned dose the thickener underflow can exceed ~45% solids, reducing tailings volume and chemical usage (research.csiro.au), and operators select polymer blends via jar testing (mdpi.com).

Rheology control spans dispersants and flocculants. Low‑molecular‑weight (low‑MW) polyanionic polymers (e.g., low‑MW polyacrylates, lignosulfonates) are added during grinding or in hydrocyclone recirculation to lower viscosity and maintain pumpable, Newtonian‑like behavior in process tanks; high‑MW polymers are added before thickening to aggregate fines for dewatering. The use of both is common; hybrid polymer formulations may enhance performance (mdpi.com). In practice, plants rely on tailored flocculants to balance settling rate and underflow density, and may deploy unit processes such as a clarifier to remove suspended solids where appropriate detention time is required.

Leaching chemistry and additives

For gold, cyanide concentration and pH are the primary variables. Typical cyanide dosage ranges from 0.5–3 kg NaCN per tonne of ore, with solution control to about 5–50 mg/L free CN. Alkalinity (lime) keeps pulp pH around 10.5–11.0 to prevent HCN gas loss and sustain Au dissolution; strict pH control aligns with Indonesia’s International Cyanide Management Code‑certified operations (cyanidecode.org). Accurate chemical dosing supports this control window, which is why plants specify equipment such as a dosing pump for cyanide or lime feeds.

Oxidants and activators expand the window for refractory feeds. Copper(II) sulfate is often dosed to activate gold occluded in copper minerals, and lead nitrate at about 50–100 g/t is used to prevent silver from cementing onto gold recovery carbon. In Indonesian research, adding an oxidant — for example, NaNO₂ at ~1.0–1.2 M — drove strong gains: one refractory concentrate achieved ~93% Au recovery with a combination of 1.2% NaCN and NaNO₂ (jurnal.tekmira.esdm.go.id), versus <50% by cyanidation alone.

For copper, sulfuric acid (often 10–30 g/L H₂SO₄) plus ferric/ferrous iron (or bacteria‑generated Fe³⁺) are standard reagents for oxide/sulfide copper; slurry pH is typically 1–2. Acid consumption tracks gangue neutralization (e.g., carbonates), and day‑to‑day auxiliaries may include lime/limestone for neutralization and antiscalants in solvent extraction circuits. Where scale control is needed in such circuits, plants consider a scale inhibitor compatible with the acid and organic systems described.

Integration, outcomes, and best practices

Energy versus recovery is the constant trade. HPGR can cut milling energy by roughly 30% while delivering several percent more metal, with economics supported by power savings and higher throughput (scielo.org.za) (ui.adsabs.harvard.edu). Ultrafine grinding can boost Au recovery in refractory ores, but at >50 kWh/t energy costs, so it is warranted when straight‑line gains (for example, +5–10% Au) outweigh energy spend. Detailed mass‑energy balances and pilot programs — as in the cited HPGR studies — should inform upgrades (scielo.org.za) (ui.adsabs.harvard.edu).

Water and reagent recycling is a quiet profit center. Operating a thickener to 45% solids (versus 30%) can cut process water recycle by approximately 30% for the same dry feed. Recovered overflow — often rich in cyanide — is pumped back to the plant to lower fresh reagent demand. Thickening tailings at 50% solids can halve dam footprint, aligning with regulations that prize minimized tailings volume and residual cyanide. Indonesian mines’ cyanide code compliance underscores this direction (cyanidecode.org).

Data‑driven control closes the loop. On‑line particle size monitors (e.g., PIDS), oxygen probes, and automated cyanide controllers hold conditions in the “ideal window.” Trending mill power, grind size, carbon loadings, and cyanide consumption supports dynamic setpoints. Case studies show that tuning cyclone cut sizes or adding dispersants in grinding can change carbon adsorption rates by 5–10%, directly impacting ounces recovered. In aggregate, combining the right equipment (e.g., HPGR), the right grind, and tailored chemistry has been documented to raise overall gold recovery by several percentage points and copper by a few percent (scielo.org.za) (ui.adsabs.harvard.edu).

Documentation and sources

Peer‑reviewed studies and industry reports underpin the guidance above (mdpi.com) (mdpi.com) (ui.adsabs.harvard.edu) (scielo.org.za) (911metallurgist.com) (research.csiro.au) (mdpi.com) (jurnal.tekmira.esdm.go.id) (cyanidecode.org). They quantify how grind size, circuit choices, and reagents move both recovery and costs.