Gold miners are rethinking oxygen: from cheap air to on‑site O₂ — and a dose of peroxide

Gold cyanidation is an oxygen‑dependent reaction: dissolved O₂ oxidizes Au⁰ to Au⁺ before complexation with CN⁻ (cyanide), and in practice available oxygen is often rate‑limiting (www.911metallurgist.com). Across modern CIL/CIP circuits (carbon‑in‑leach/carbon‑in‑pulp), operators therefore add oxygen via compressed air sparging, direct O₂ from on‑site generators (PSA/VSA; pressure/vacuum swing adsorption), cryogenic plants, or supplemental chemical oxidants such as hydrogen peroxide (H₂O₂). Each route trades off oxygen transfer efficiency, cost, and operating impact — and the differences are not subtle (www.researchgate.net).

Compressed air sparging limits

The baseline method is compressed air: cheap gas, but only ~21% O₂ by volume. Even with vigorous agitation, air‑sparged pulp rarely exceeds ≈8 mg/L dissolved oxygen — the saturation at 1 atm O₂ — and in practice measured values are much lower, about 0.5–1.7 mg/L in an air‑agitated bottle (www.researchgate.net). Plant data capture the inefficiency: an O₂ “oxidation efficiency” of ≈0.6% for air versus ≈8.6% when pure O₂ is used (www.researchgate.net).

High air flow rates are often needed to compensate, driving up compressor power and creating foam, dusting, and blower maintenance. Many mines find air sparging cannot meet the true oxygen demand of refractory or viscous pulps without extremely long retention times (www.911metallurgist.com; www.researchgate.net).

On‑site oxygen generation (PSA/VSA)

On‑site O₂ plants (PSA/VPSA) deliver 90–95% purity oxygen at controlled flow. Industry experience shows a step‑change versus air: replacing air with liquid O₂ injection in a CIL tank cut the tailings Au grade from 1.91% to 1.69%, reflecting higher metal dissolution (www.researchgate.net). Dissolved O₂ rose from ~0.5–1.7 mg/L under air to >20 mg/L with pure O₂ — far above the 8.2 mg/L saturation benchmark — accelerating leach kinetics and boosting recovery (www.researchgate.net).

PSA/VPSA units operate at ambient temperature with low‑pressure air over zeolite beds, delivering continuous ~90–95% O₂ without cryogenic cooling. With low specific energy consumption and modular construction, they are cost‑competitive at gold‑mine scales (10–300 tonnes O₂/day) and are “proven ideal for gold leaching,” enabling “improved gold recoveries, higher plant throughput, [and] greater energy efficiency” (www.afrox.co.za; www.afrox.co.za).

Press reports corroborate large benefits: on‑site PSA oxygen has “yielded faster leaching, greater recovery and less reliance on toxic chemicals,” and mines switching from delivered to on‑site oxygen report up to 70% cost savings (www.engineeringnews.co.za; www.engineeringnews.co.za). Deploying on‑site oxygen generators has been shown to improve gold output while cutting cyanide use and carbon footprint, and also eliminates delivery delays and associated emissions (www.engineeringnews.co.za; www.engineeringnews.co.za). Many sub‑Saharan mines now adopt on‑site PSA and even combine it with supplemental cylinder/lunch deliveries at peak demand, reducing reliance on a single supplier (www.engineeringnews.co.za; www.engineeringnews.co.za).

Cryogenic oxygen supply

Cryogenic plants (ASUs) produce near‑100% O₂ via low‑temperature distillation at high pressure. The method is energy‑intensive and best where very large, continuous flows of ultra‑pure O₂ (>99%) are needed. Most gold operations find 90–95% O₂ sufficient and therefore favor PSA/VPSA; for remote gold plants, cryogenic capital and transport logistics usually outweigh benefits.

Chemical oxidants (hydrogen peroxide)

Chemical oxidants are used as a supplemental oxygen source when gas transfer is difficult. Hydrogen peroxide (H₂O₂) decomposes catalytically on ore surfaces to yield dissolved O₂ in situ, entirely avoiding gas–liquid transfer limits (www.911metallurgist.com). Typical working dilutions are 0.5–5% w/w, and dosing can surpass normal air saturation (~8 mg/L) by >100% “if this is required,” whereas compressed air is “significantly less effective” at raising DO (www.911metallurgist.com; www.911metallurgist.com). To deliver the same active O₂ by air would require ~800× the gas volume compared with H₂O₂ (www.911metallurgist.com).

Recovery gains tend to be modest. Heap tests on a Kazakh low‑sulfide ore showed +1.2 percentage points gold recovery with 0.5% H₂O₂ (55.89% vs 54.69% without), and bottle rolls saw 52.92% Au vs 52.03% without (www.researchgate.net). A drawback is cyanide burn: H₂O₂ can oxidize CN⁻ to cyanate (NaCN + H₂O₂ → NaOCN + H₂O), often causing a small rise in cyanide consumption — about +0.04 kg NaCN/t in the Kazakhstan study (www.911metallurgist.com; www.researchgate.net). Degussa’s guidance is to restrict H₂O₂ to <5% and use feedback O₂ electrodes to avoid “detoxification” (www.911metallurgist.com). In this context, mines procure oxidants as part of a broader chemicals program for mining (Chemicals for Mining applications).

Cost and practicality are the catch: reagent‑grade 50% H₂O₂ often costs on the order of USD 400–700/ton at mine sites (varies widely) and is more expensive than compressed air as an oxygen supplier, though higher gold output can justify the cost. It breaks down to water and oxygen without gas emissions but requires careful handling, storage stability, and onsite tanks; operators typically dose intermittently or in a pump‑around reactor (www.911metallurgist.com). Where tight control is required to minimize CN burn, accurate chemical dosing is a key control lever (dosing pump).

Alternatives are uncommon in‑core: ozone (O₃) must be generated on‑site and is expensive at scale; hypochlorite or chlorine has historical precedent but is corrosive/toxic and largely replaced by H₂O₂ or pressure oxidation. In copper acid leaching, strong oxidants (ferric sulfate or oxygen in steam–oxygen mixes) are used for chalcopyrite, but those high‑temperature/pressure routes (e.g., Cu POX) sit beyond typical heap/CIL setups.

Measured outcomes and kinetics

Under typical conditions (~1 atm total pressure), even 90% O₂ raises the oxygen partial pressure by ~4.5× compared to air (20% O₂), and Henry’s law then boosts dissolved O₂. Plants injecting oxygen often report 3–10% higher absolute Au recovery on refractory ores and shorter leach cycles by 20–50% (percentages vary by ore), with lower cyanide usage in well‑oxygenated CIL tanks because more CN⁻ goes to Au complexation rather than side‑reactions (www.researchgate.net; www.911metallurgist.com). On a pyrrhotite‑rich ore, pre‑oxygenation to a (CN⁻×O₂) index of ≈5.6 mmol²/L enabled 83% Au extraction in the first leach tank, underscoring how abundant O₂ can shorten retention times (www.researchgate.net).

Regulatory context (Indonesia)

Indonesia enforces strict cyanide management — for example, PT J Resources Nusantara’s Bakan mine is recertified under the International Cyanide Management Code — but there are no special prohibitions on oxygen technologies. Standard industrial safety and environmental regulations apply (e.g., boiler/pressure vessel codes for oxygen plants; discharge limits for tailings) (cyanidecode.org). The Ministry of Environment sets limits on metals (As, CN, etc.) in discharge, and larger O₂ plants may require electrical and construction permits under mining regulations; these processes are routine (www.minesandcommunities.org).

Illustrative performance metrics

Values vary by plant and ore. Representative metrics and their sources are summarized below.

• Compressed air: ~21% O₂ (N₂ balance); typical dissolved O₂ ~0.5–2 mg/L with a practical ceiling near ~8 mg/L saturation; O₂ oxidation efficiency ≈0.6% vs ≈8.6% for pure O₂; low gas cost but high blower power; foam/dusting and maintenance issues; often cannot meet oxygen demand of refractory or viscous pulps without very long retention times (www.researchgate.net; www.researchgate.net; www.911metallurgist.com).
• On‑site O₂ (PSA/VPSA): ~90–95% O₂; dissolved O₂ often >20 mg/L at 1 atm; reported gains of +3–10 percentage points Au recovery and 20–50% shorter leach; lower cyanide use in well‑oxygenated tanks; higher capex but low operating cost per ton O₂; up to ~70% savings vs delivered oxygen; modules cover ~10–300 t/d capacity and improve throughput and energy efficiency (www.researchgate.net; www.engineeringnews.co.za; www.afrox.co.za; www.afrox.co.za; www.engineeringnews.co.za).
• Cryogenic O₂: ~99% O₂ (liquid/high pressure); similar high DO at 1 atm; suitable for very large, continuous flows; high capital and energy for plants; often delivered in tanks otherwise.
• Hydrogen peroxide (0.5–5%): 100% O₂ equivalent delivered in solution; DO can greatly exceed ~8 mg/L saturation; typical gold recovery uplift on tough ores is modest (+1–2 percentage points in reported trials); small increase in cyanide consumption (~+0.04 kg NaCN/t reported); reagent cost high (hundreds of dollars per ton reagent); used intermittently; leaves no gas emissions but demands careful handling/storage (www.911metallurgist.com; www.researchgate.net; www.911metallurgist.com; www.911metallurgist.com).

Bottom line

Compressed air remains ubiquitous but limited. Pure oxygen — especially via on‑site PSA/VPSA — consistently elevates dissolved O₂, shortens leach times, and raises recovery, with logistics and operating cost advantages over delivered O₂. Hydrogen peroxide is a potent supplemental oxygen source for difficult circuits, fully dissolving as O₂ in pulp, but its benefit is modest and must be weighed against reagent expense and cyanide burn (www.911metallurgist.com; www.researchgate.net; www.engineeringnews.co.za).