The quiet tweaks slashing organic losses in copper SX

In solvent extraction (SX) circuits for copper and gold, every misplaced droplet of organic diluent is money leaving the plant. The fixes aren’t flashy: slightly larger settlers, full‑width baffles, a coalescer pad here, a short flotation step there, and—when needed—measured shots of surfactant or salt. But the numbers are decisive: designing for 1–2 minutes of calm coalescence, then polishing the discharge, consistently cuts organic carryover from the hundreds of ppm to single digits (911metallurgist.com; patents.google.com).

What’s changing is precision. Designers are pinning down settler flux, dispersion‑band height, and inlet hydraulics—and backing it with after‑settlers, coalescers, or flotation—so each stage achieves clean phase split without heroic chemistry (scielo.org.za; 911metallurgist.com).

Settler flux and residence time targets

Effective phase separation starts with adequate settler capacity. Designers target a specific settler flux (flow per area) low enough for minute‑scale dwell times. For copper SX, values around 2–4 US gpm/ft² (~8–10 m³/m²·h) are typical (www.scielo.org.za). At 2.4 gpm/ft² with a 4″ (≈10 cm) dispersion‑band in the settler, the mixed phase resides ~1 min in the band (911metallurgist.com).

Benchmarks for the fluids themselves are equally tight: fresh SX organics typically need ~30–60 s to disengage, while spent or contaminated organics can take up to ~2 min (patents.google.com). That’s why a 1–2 min total residence (mixer plus settler) is often required for near‑equilibrium extraction. Hydraulics matter: higher downcomer velocities (e.g., shallow organic height) shorten this time dramatically. A 32 m‑wide settler at 1000 m³/h with only 0.2 m organic height gives a linear velocity ~312 m/h and only ~3.9 min through a 20 m length—clearly insufficient (www.scielo.org.za).

The takeaway: settler area and organic bed depth should be sized to allow ≥1–2 min of calm coalescence for the expected flow, yielding >90% phase clarity. In practice, designing for a 4–6″ dispersion band and ≳1–2 min residence per stage keeps residual organic entrainment to very low (<20–50 ppm) levels (911metallurgist.com; patents.google.com).

Inlet and outlet baffling strategy

Key vessel features close the gap between design and reality. “Picket‑fence” baffles spanning the full weir width spread the emulsion uniformly and break up dispersion layers, dramatically reducing entrainment (patents.google.com). Modern practice notes that full‑width inlet/outlet baffles are “the primary means of controlling entrainment” when flow is moderate (patents.google.com).

Inlet distributor baffles ensure the mixer discharge fills the settler cross‑section (911metallurgist.com; patents.google.com). These internals, along with maintaining the fluid velocity within recommended limits, keep O‑in‑A (organic‑in‑aqueous) losses to tens of ppm. Equipment guidelines often recommend oversizing final‑stage settlers (extra area or depth): small amounts of entrainment in earlier stages have negligible effect, but final‑stage carryover directly costs organic losses (911metallurgist.com).

Polishing stages: coalescers and flotation

Even with ideal settler sizing, trace organic remains. Most plants use polishing stages to scrub out entrained solvent. A common approach is a two‑step cascade: pump the aqueous phase—raffinate or loaded electrolyte (the metal‑bearing aqueous stream leaving extraction or stripping)—through a quiet surge tank or coalescing chamber, then through engineered filters or flotation. One design sends loaded electrolyte from the stripper first to a surge tank, then to a solvent‑recovery section with either coalescing packs or a flotation cell to remove residual organics (911metallurgist.com).

Coalescing media (“de‑misters”) are often used. Nonwoven polypropylene or fiberglass pads (pleated or stacked) can be inserted at settler outlets or in side‑stream filters. TREA (3M) highlights that a pleated nonwoven coalescer placed near the settler overflow can capture very fine droplets (e.g., <30–40 µm) and “provide disruption of the interfacial film,” yielding much lower entrainment in both phases (trea.com). In practice, coalescers have demonstrated 90+% removal of micro‑droplets: passing a 50–100 ppm organics‑laden stream through a static coalescer often reduces outlet levels to single ppm. (Care: such packs can clog if solids are present.) Where solids are part of the duty, engineered filters—such as cartridge elements installed in corrosion‑resistant housings—are the typical companion; in SX environments, stainless housings like stainless steel cartridge housings and upstream cartridge filters are commonly selected to manage particulate load before the coalescer.

Where solids or suspended crud make filters impractical, flotation is an alternative. Electrolyte or raffinate is aerated in a flotation column, causing organic droplets to rise and overflow. Literature reports very high recovery: >90% of organics can be removed even at high loading. One source notes that an optimum 2.5 min flotation cleared ~90% of organics from streams initially at 50–100 ppm, and longer times add only marginal benefit. Flotation is effective even at 1000 ppm; only slightly longer contact is needed (911metallurgist.com). Plants implementing this step often favor packaged dissolved‑air designs for convenience, for instance a dissolved air flotation unit.

Auxiliary columns and after‑settlers

Some plants use dedicated coalescing columns on critical streams. The Chuquicamata SX plant employs an auxiliary column packed with fibrous media on the loaded‑organic line to ensure any trapped aqueous droplets are removed before recycling (www.scielo.org.za). Similarly, “after‑settler” tanks or columns can be placed on raffinate discharge—essentially small gravity separators (sometimes with internals or packing) that capture any lingering organic.

The combination of these devices—picket fences, coalescer packs, flotation cells—typically reduces organic losses from the hundreds of ppm down to single‑digit ppm in the outlet water (911metallurgist.com; 911metallurgist.com).

Chemical additives for disengagement

Chemical agents can accelerate phase split. Most are surfactant‑ or salt‑based demulsifiers added upstream of settling. Nonionic surfactants—such as polyethylene–polypropylene glycol block copolymers—have been identified as effective “phase disengagement agents.” Adding on the order of 100–500 ppm of a tailored nonionic surfactant to the aqueous feed can cut the phase disengagement time by ~25–30% (e.g., reducing a 100 s separation to ~75 s) (patents.google.com). In practical terms, jets of demulsifier into the feed tank or mixer can clear emulsion films faster, limiting carryover; these patents emphasize non‑ionic, moderate HLB surfactants and exclude thiocarbonyl compounds to avoid stability issues.

Metered addition is standard practice in process control; plants typically dose such chemistries via metering equipment like a dosing pump so the ppm‑level targets are held steady at the mixer or feed tank.

Another novel additive is water‑soluble metal salts. BASF/Cognis has patented adding small doses of “compatible salts” (e.g., CuSO₄, FeSO₄, Al₂(SO₄)₃) directly into the organic or mixed emulsion. The idea: salt crystals nucleate or charge‑balance the fine suspended droplets, forcing rapid coalescence. In copper SX practice, the recommendation is only ~1–2 kg of CuSO₄ per 50 tons of Cu produced (on the order of 100–200 ppm a few times per day) (patents.google.com; patents.google.com). Trials indicate that spraying fine CuSO₄ crystals into the mixer/emulsion makes the dispersion “collapse” faster in the settler, at the cost of introducing solid sulfate sludge (“crud”) that must be removed.

Other additives include selective demulsifiers or flocculants. Trace amounts of polyvalent cation sources (Ca²⁺, Mg²⁺) or filtration aids (bentonite, silica) are sometimes used to agglomerate colloidal impurities and break thin films. In practice, operators may also “polish” recirculating organics by clay filtration to strip out fine crud that acts as emulsifier—though clay often creates a separate disposal issue. Compatibility is critical: e.g., using Ca or Al salts only if they don’t introduce unwanted metals or change chemistry adversely (patents.google.com).

Measured impact and plant data points

Measured improvements are material. One SX circuit reported that pre‑treating organic with 0.5% bentonite (removing fines) reduced aqueous entrainment on the organic by ~40–65%, roughly halving the impurity load (scribd.com). In the broader flowsheet, the combination of conservative settler flux (2–4 US gpm/ft²), controlled dispersion bands (4–6″), targeted baffling, and a polishing step (coalescer or ~2.5 min flotation) is what consistently drives entrainment from tens to single‑digit ppm (scielo.org.za; 911metallurgist.com).

For operators standardizing hardware around these controls, supportive plant items—like corrosion‑resistant steel filter housings for side‑stream polishing or an on‑skid DAF package for flotation duty—are the practical enablers. But the core levers remain the same: conservative settler design, full‑width baffling, simple polishing steps, and judicious chemistry (patents.google.com; 911metallurgist.com).

References and sources (all cited inline): settler and coalescer data from metallurgical design studies (911metallurgist.com; patents.google.com; patents.google.com) and case studies (scielo.org.za; scielo.org.za). Coalescer/flotation performance from industry sources (911metallurgist.com; 911metallurgist.com). Additive roles from recent patent literature (patents.google.com; patents.google.com). Each shows how design tweaks and targeted chemistry produce quantifiable drops in organic carryover (often reducing entrainment from tens of ppm to low ppm or less).