Solvent Loss in SX: 7 Causes of Organic Inventory Loss

Last updated: 2026-11-08

Most solvent loss in a solvent extraction plant is invisible: a mid-size copper SX circuit can shed 30–80 L of organic per hour as raffinate entrainment, crud, and carryover — 300–600 m³ per year — without a single visible drop on the ground. This article maps where the organic inventory goes, how to audit it, and which fixes pay back first.

Key Takeaways

• Entrainment of organic in raffinate is the largest single cause of solvent loss, typically 20–150 ppm of raffinate flow.
• Skimmed crud quietly exports 30–60 wt% organic in every discarded interfacial band.
• An organic inventory balance — dam levels, make-up, raffinate TOC — closes the loss split in six steps.
• Hot-climate plants lose 0.5–2.0 L per m² of open settler surface per month to evaporation.
• Launders, coalescers, and crud distillation recover 60–90% of lost organic and typically pay back in 2–4 years.

What Is Solvent Loss? Definitions and Loss Pathways

Solvent loss is any pathway by which the organic phase — diluent, extractant, and modifier together — permanently leaves the solvent extraction circuit that was designed to recirculate it for years. In copper SX the organic is working capital: purchased once, conditioned over months, and expected to circulate indefinitely. A plant that loses organic pays twice, once to replace the reagent and again to clean the aqueous streams it contaminated.

Four definitions frame the rest of this article:

Entrainment is the carrying of fine droplets of one liquid phase inside the other phase that exits a mixer-settler. Organic-in-aqueous entrainment in raffinate is the dominant solvent loss pathway in most plants; aqueous-in-organic entrainment is its mirror image and behaves very differently, so the two are treated separately throughout.

Crud is a stabilized interfacial emulsion of fine solids — silica, iron and aluminum hydroxides, manganese dioxide — that adsorbs extractant and diluent into a viscous band riding the aqueous–organic interface. Every cubic meter of crud skimmed and discarded exports its organic fraction with the solids.

An organic inventory balance is a mass accounting exercise that compares purchased make-up against losses measured at every outlet — raffinate, loaded and strip liquor, crud, evaporation, leaks, and degradation — so the split can be ranked and attacked in order of size.

Adsorption loss is the consumption of dissolved extractant and diluent onto suspended solids that leave the circuit in aqueous streams, a quieter cousin of crud that never shows up at the interface.

The scale of the problem is easy to underestimate. The organic circuit of a mid-size copper plant holds 500–1,500 m³ of organic (typical range, confirm with test work), so small percentages compound: a circuit losing only 0.1% of its inventory per day turns over its entire stock every 2.7 years — and most losing plants lose considerably more than that.

Key Parameters: How Do You Detect Solvent Loss Early?

Detection is a sampling discipline, not a single instrument. Total organic carbon (TOC) analyzers report milligrams of carbon per liter and are the fastest routine screen for raffinate quality; gas chromatography with flame ionization detection (GC-FID) resolves the hydrocarbon species and reports true organic phase content in ppm by volume. A weekly composite sample from the raffinate launder, the advance electrolyte, and the strip liquor, converted through each stream’s flow rate, turns ppm into liters per hour — and liters per hour into a maintenance budget. The conversion is direct: a raffinate flowing at 750 m³/h carrying 40 ppm of organic is exporting 30 L/h, every operating hour.

Visual and volumetric indicators fill the gaps between lab results. A rainbow sheen at the raffinate weir or on the pond means entrainment above roughly 30 ppm at that instant. Organic tank and dam levels, reconciled against metered make-up, expose losses that no inline analyzer will catch — evaporation and spillage in particular. Make-up purchase records are themselves a detection tool: when diluent consumption rises steadily over months while production holds flat, something in the circuit is consuming inventory.

Table 1 summarizes the loss mechanisms, their typical magnitudes, and how each is detected (all values are typical ranges — confirm with test work on your own plant).

Two readings of that table matter for day-to-day control. First, a single raffinate result is noise; a monthly trend is signal, so control limits should be set on rolling averages — alert at 50 ppm, act at 80 ppm (indicative, set from your own baseline). Second, the mechanisms with no direct analyzer — evaporation, leaks, degradation — are exactly the ones the inventory balance residual exposes, which is why the balance is the backbone of any serious audit.

Where Does the Organic Inventory Go? The Seven Causes of Solvent Loss

Every losing plant we have audited traced its deficit to one or more of the seven mechanisms below. Each is presented as observed symptom, root cause, then corrective action.

1. Organic entrainment in raffinate. Symptom: steady organic make-up, a rainbow sheen on the raffinate pond and downstream dams, and GC-FID results of 40–150 ppm against a target below 50 ppm. Cause: organic droplets too fine to coalesce within the available settler residence time. The usual drivers are excessive mixer shear, a dispersion band run too deep, insufficient settler area for the duty, and suspended solids or surfactants stabilizing the droplets against coalescence — the whole subject is treated in our SX entrainment note. Fix: reduce mixer tip speed and impeller submergence, control the dispersion depth with weirs and level control, extend residence time or area, and install full-flow organic launders on the raffinate exits. Phase chemistry matters too; poor phase disengagement multiplies every hydraulic defect.
2. Organic leaving with loaded and strip liquors. Symptom: sheen and floating globules in the EW advance electrolyte, pitted or “burnt” cathode patches, and organic accumulating in tankhouse bleed streams. Cause: organic carryover from the loaded settler rides forward into electrowinning, where there is no recycle path back to SX — the electrolyte is not returned, so this organic is lost to the circuit and must be skimmed from cells or it contaminates cathode. Fix: launders and coalescers on the loaded liquor exit, a maintained phase boundary in the loaded settler, and skimming or floatation on the EW electrolyte storage.
3. Crud adsorbing organic, and adsorption onto suspended solids. Symptom: a thickening interfacial band, more frequent skimming campaigns, make-up rising with no visible aqueous loss, and discarded crud that releases 30–60 wt% of its mass as organic when distilled. Cause: fine solids — silica leached from ore, iron and aluminum hydroxides from pH excursions, manganese dioxide from hot acid circuits — stabilize the interface into an emulsion that traps organic; the same solids, leaving with aqueous streams, carry adsorbed reagent out quietly. Fix: tighten leach clarification, control pH and silica input, and run a proper crud treatment loop that recovers organic instead of landfilling it — the options are compared in our crud treatment page.
4. Aqueous-in-organic carryover returning in reverse. Symptom: organic make-up climbing while raffinate TOC stays flat; water pooling in organic storage tanks; an unstable, wandering interface in the strip settler. Cause: aqueous droplets carried forward inside the organic phase to the strip stage, where they finally settle and leave with strip-bleed and wash streams — carrying dissolved and re-entrained organic out of the loop in a direction nobody samples. Fix: correct mixer O:A settings and geometry so the organic never slugs forward; the set-point logic is on our O:A ratio page.
5. Evaporative and VOC loss from open settlers. Symptom: losses worst in the hot dry season, dam levels falling faster than metered streams explain, and no sheen anywhere to blame. Cause: kerosene-based diluent evaporates from hot open surfaces, and wind strips the vapor layer; the loss scales with exposed surface area, temperature, and wind speed. Fix: quantify first via the balance residual, then cover settlers, float Balls or boards on quiescent zones, and erect windbreaks where covers are impractical.
6. Physical leaks and spillage. Symptom: stained pump bases, hydrocarbon odor in bunds, sudden step-changes in organic dam levels, soil staining along pipe racks. Cause: worn mechanical seals, hose failures, overfilling during make-up campaigns, and bund drains left open. Fix: weekly walk-downs with a checklist, intact bunding and spill kits, level alarms on organic storage, and a make-up procedure that records every liter added.
7. Chemical degradation consuming reagent. Symptom: extractant strength in the organic phase drifting down with no organic found anywhere in the aqueous outlets, and reagent additions needed just to hold the extraction profile. Cause: extractant molecules oxidize or hydrolyze over time — accelerated by heat, chloride and nitrate in feed solutions, dissolved oxygen, and copper-catalyzed side reactions — and the degradation products are unrecoverable. Fix: track extractant strength by titration against a fresh-organic standard, control temperature and impurity ingress, and review the chemistry in our extractant degradation note.

How Do You Audit Solvent Loss in an SX Plant?

The audit below closes an organic inventory balance in six steps and typically needs four to six weeks of data. Run it monthly once established; the first pass usually lands the solvent loss split within ±15%, which is accurate enough to rank causes and fund fixes.

1. Baseline the organic inventory. Survey every organic tank, dam, sump, and settler heel with a strapping or dip measurement, convert levels to liters using verified vessel dimensions, and record each figure with a date and time. This snapshot is the balance’s opening stock; without it, nothing downstream reconciles.
2. Rebuild the make-up ledger. Sum every diluent and extractant delivery since the last audit from purchase records and tank farm receipts, subtract any organic sent off-site for disposal or return, and correct for measured density so drums and bulk loads are both in liters. Declare the accounting basis — the spec point matters when drums are stored half-empty around the plant.
3. Assay every aqueous outlet. Take weekly composite samples of raffinate, loaded liquor, advance electrolyte, and strip bleed; analyze by GC-FID where available, TOC otherwise, and convert concentration to loss rate as flow × ppm. Report raffinate TOC in the 10–80 mg/L band as an alert zone (typical range, confirm with test work); a raffinate carrying more than ~100 mg/L of TOC is exporting organic faster than most plants realize.
4. Account for crud and solids-bound organic. Record the volume of crud skimmed per week and measure its organic content — distillation of a composite sample is the defensible method, since crud holds 30–60 wt% organic typically. Multiply volume by organic fraction to get liters exported per week, and log it separately from aqueous entrainment so the two causes stay distinguishable in the ranking.
5. Estimate the unmeasured losses. Walk the plant weekly for leaks and spillage and price them in liters; estimate evaporation from open surface area and climate. The balance residual — make-up minus all measured outlets — then attributes what remains to evaporation, spills, and degradation combined; if the residual exceeds ~15% of make-up, suspect a leak or an unsampled outlet before blaming chemistry.
6. Close the balance and rank. Express every stream as a percentage of total make-up, present the split in a single table, and fix the top two causes first — they are usually more than 70% of the deficit. Set standing KPIs: raffinate ppm alert level, monthly balance residual, crud organic recovery rate, and make-up liters per tonne of cathode.

Which Design Levers Cut Solvent Loss at the Source?

Fixes divide into hydraulic, interfacial, and chemical, and the sequence matters: hydraulic corrections come first because they attack the largest stream of solvent loss and are the cheapest to install. Settler design is the dominant lever — residence time of 3–8 minutes with shallow, uniform flow lets droplets settle rather than export, and launder geometry on every aqueous exit captures what does not. Mixer adjustments — lower tip speed, deeper impeller submergence, correct O:A ratio — reduce the droplet population at its source; the equipment side is covered in our mixer-settler design overview, and construction material matters in hot acid service, where polypropylene mixer-settlers and stainless steel mixer-settlers each carry corrosion and cost trade-offs.

The sequencing rule we apply on projects: install launders and fix mixer hydraulics first — they address the largest stream and often halve the deficit within a quarter; add coalescers or carbon polishing only after measuring what survives the first round; and treat crud distillation as a parallel workstream because its economics stand alone. Any pilot-scale confirmation should reproduce your actual feed solids and phase chemistry, since entrainment behavior scales with chemistry as much as with geometry.

What Does Solvent Loss Cost per Liter?

Put a price on the inventory before you prioritize anything. Declared assumptions (indicative global bands, confirm with current supplier quotes): kerosene-based diluent at $0.85–1.00 per liter and extractant at $5–8 per liter. A copper SX organic blended at 6 vol% extractant then costs roughly:

0.94 × $0.90/L + 0.06 × $6.00/L ≈ $0.85 + $0.36 ≈ $1.21 per liter

A circuit losing 50 L/h over 8,000 operating hours per year sheds 400 m³ of organic annually — about $480,000 per year at the blended price, before any downstream damage. And replacement is only the visible half of the bill. Organic reaching the EW tankhouse pits and burns cathode and pulls product toward the edge of specification; organic leaving in raffinate runs into environmental permitting, where hydrocarbon and TOC limits on effluent are tightening across most jurisdictions, and a discharge excursion is a compliance event, not a reagent cost. Add the hidden charges — crud disposal fees that include someone else’s organic, extinguished reagent value in degraded extractant, and labor for endless skimming — and the effective cost of solvent loss typically lands 1.5–2 times the blended price per liter. That multiplier is why recovery hardware, not cheaper reagent purchasing, is almost always the better spend.

A Worked Case: Tracing a 50 L/h Solvent Loss (Synthetic)

The following case is synthetic and illustrative, assembled from typical copper SX practice rather than any specific plant; the copper industry flowsheet it assumes is standard. The plant produces 120,000 t/y of cathode, operates 8,000 h/y, and pumps 750 m³/h of raffinate. A three-month organic inventory balance showed make-up running 50 L/h above production needs — a solvent loss rate worth nearly half a million dollars a year.

The split closed as follows. Raffinate entrainment: 750 m³/h × 40 ppm = 30.0 L/h, or 60% of the deficit. Loaded and strip liquor carryover: 300 m³/h × 15 ppm = 4.5 L/h, or 9%. Crud export: 2 m³/wk skimmed at 45 wt% organic ≈ 0.9 m³/wk ≈ 5.4 L/h, or 11%. Evaporation from open settlers: ~2 L/h, or 4%. Leaks and spillage: ~3 L/h, or 6%. Degradation plus balance residual: ~5.1 L/h, or 10%. Sum: 30.0 + 4.5 + 5.4 + 2.0 + 3.0 + 5.1 = 50.0 L/h — the balance closed exactly, at 400 m³/y and roughly $480,000 per year at $1.20/L blended.

Two fixes were funded first. Full-flow launders on the raffinate exits recovered 75% of that stream, cutting it from 30.0 to 7.5 L/h net. A crud distillation unit recovered 60% of crud-held organic, cutting 5.4 to 2.2 L/h net. The revised deficit: 7.5 + 4.5 + 2.2 + 2.0 + 3.0 + 5.1 = 24.3 L/h, or about 194 m³/y — roughly $233,000 per year. Against a combined capital cost of about $700,000 for the two projects, the payback was approximately 2.8 years, with reduced EW contamination as an unpriced bonus. The ranking, not the hardware, was the insight: without the balance, the plant would have spent the same capital on activated carbon and missed 60% of the problem.

Frequently Asked Questions

How much solvent loss is normal in a copper SX plant?

Well-run circuits hold raffinate entrainment below 50 ppm and total make-up below roughly 0.05% of circuit inventory per day (typical ranges, confirm with test work). A plant above 100 ppm raffinate organic, or whose organic make-up exceeds 0.1% of inventory daily, has a measurable problem worth an audit. The right benchmark is your own rolling balance, not a neighboring plant’s, because feed solids and climate dominate the split.

What is the difference between entrainment and aqueous-in-organic carryover?

Entrainment is fine droplets of one phase carried inside the other: organic droplets in raffinate (organic leaves the circuit) or aqueous droplets in organic (water moves forward, destabilizing downstream settlers). Carryover of aqueous inside organic is the reverse-facing problem — it rarely loses organic directly, but the aqueous that finally separates in strip and wash circuits carries dissolved and re-entrained organic out of the loop, and it degrades strip performance first.

How do I measure organic in raffinate — TOC or GC-FID?

Use both, for different jobs. GC-FID resolves hydrocarbon species and reports true organic content in ppm v/v, which converts directly to liters per hour; it is the audit-grade method. TOC is cheaper and faster, reports milligrams of carbon per liter, and is ideal for weekly trend and alert monitoring. Calibrate TOC against GC-FID on your own raffinate once, then run TOC routinely with periodic GC-FID checks.

Can organic be recovered from crud?

Yes, and economics usually support it. Skimmed crud typically holds 30–60 wt% organic (typical range, confirm with test work), and distillation or acid-plus-diluent washing recovers 50–80% of it. The recovered organic returns to circuit at full value, so a crud distillation unit on a plant skimming 2 m³/wk recovers roughly 100 L/wk — meaningful money at blended organic prices. The remaining solids still require disposal, now as a smaller, de-oiled stream.

Why does solvent loss contaminate the electrowinning product?

Organic carried forward with loaded liquor reaches the EW cells, where it floats on the electrolyte, wets the cathode surface, and locally blocks metal deposition — producing pitted, dark-edged deposits that can drift out of specification. Organic levels above roughly 5 ppm in advance electrolyte are visible as surface sheen, and even lower levels can mark cathode over long cycles. Launders on the loaded settler and skimming on electrolyte storage are the standard defense.

Do launders pay for themselves?

Usually, and quickly. A full-flow launder recovers 60–90% of entrained organic at essentially zero operating cost — it is passive hardware. On a plant losing 30 L/h to raffinate entrainment, a 75% recovery returns about 22.5 L/h, worth roughly $220,000 per year at a $1.20/L blended price (assumptions declared in the cost section). Launders typically pay back in one to three years, faster where raffinate flows and entrainment levels are both high.

How does organic degrade, and can degradation losses be prevented?

Extractant molecules slowly oxidize and hydrolyze, accelerated by high temperature, chloride and nitrate in feed, dissolved oxygen, and copper-catalyzed reactions; degradation products hold no extraction value and are unrecoverable. You cannot stop the chemistry, but you can slow it: control settler and mixer temperatures, limit chloride and nitrate ingress, avoid air entrainment in the organic, and track extractant strength by titration so degradation is separated from physical loss in your balance.

Conclusion

Solvent loss is rarely one leak; it is a split across seven outlets, and the two largest — raffinate entrainment and crud export — usually account for more than 70% of the deficit. The path to control is the same everywhere: run a monthly organic inventory balance on dam levels, make-up records, and raffinate TOC, let the residual expose the losses no analyzer can see, then spend capital on launders, coalescers, and crud recovery in order of size rather than habit. A plant that closes its balance typically recovers half its organic make-up within a quarter, and most recovery hardware pays for itself inside three years while relieving environmental permitting pressure downstream. If you are auditing an existing circuit, sizing a new SX train, or choosing between launder and coalescer options, request a quote with your raffinate flow, current make-up figures, and target ppm — our process team will review your mixer-settler configuration against your own numbers.

References

1. J. Rydberg, C. Musikas, G.R. Choppin (eds.), Principles and Practices of Solvent Extraction, Marcel Dekker, New York.
2. G.M. Ritcey, Solvent Extraction — Principles and Applications to Process Metallurgy, revised 2nd edition, G.M. Ritcey & Associates, Ottawa.
3. F. Habashi, A Textbook of Hydrometallurgy, Métallurgie Extractive Québec, Québec City.
4. Society for Mining, Metallurgy & Exploration (SME), reference literature on extractive metallurgy, copper hydrometallurgy sections.
5. TMS and MetSoc of CIM conference proceedings, papers on copper SX crud formation, organic entrainment, and solvent recovery practice.
6. Hydrometallurgy (Elsevier) journal literature on crud mechanisms, extractant degradation, and organic entrainment in copper solvent extraction — cited as general industry literature, no public URL.

Changelog

• 2026-11-08: First published.