Mixer Settler Design: Sizing, Stages & Selection Guide

Last updated: 2026-10-21

A sound mixer settler design typically uses 2–3 minutes of residence time per mixing compartment and a settler sized for roughly 3–8 m³/h of total phase flow per m² of settling area in copper SX duty. Getting these two numbers right fixes most of the plant’s extraction performance, entrainment behavior, and capital cost before detailed engineering even begins.

This guide is part of our liquid-liquid separation hub, where we cover the full solvent extraction equipment family. It is written for process engineers and metallurgists who need to size, specify, or select a mixer-settler battery for hydrometallurgical duty — most commonly copper solvent extraction, but the same logic applies to nickel, cobalt, zinc, and rare-earth circuits.

Key Takeaways

• Mixer settler design starts with residence time: 2–3 minutes per mix stage is typical for copper oxime systems.
• Settler area is set by total phase flow per unit area, typically verified with bench and pilot disengagement tests.
• Most copper circuits need 2 extraction stages plus 1 or 2 strip stages for acceptable recovery and acid transfer.
• Impeller tip speed below about 3.5 m/s limits shear, crud formation, and organic entrainment in aqueous raffinate.
• Stage efficiency of 85–95% is achievable with well-designed mixing; pilot testing confirms the design margin.

What Is a Mixer-Settler and How Does It Work?

A mixer-settler is a stage-wise liquid-liquid contactor that mixes two immiscible phases — an aqueous phase and an organic phase — and then allows them to disengage by gravity in a separate quiescent zone. A mixer-settler is a mass transfer device that repeatedly contacts, equilibrates, and separates the two phases so that a solute (for example, copper ions) transfers from the aqueous phase (pregnant leach solution, PLS) into the organic phase (a diluent carrying an extractant). A mixer-settler is also a hydraulic machine whose pump-mix impeller must both blend the phases and transfer the combined flow to the next stage, which is why its design couples chemistry, fluid mechanics, and tank geometry in a way that columns and centrifugal contactors do not.

Solvent Extraction: selection rules

In operation, aqueous feed and organic enter a mixing compartment where an impeller disperses one phase as fine droplets into the other. Interfacial area — not residence time alone — governs extraction kinetics, so the impeller must create enough droplet surface for mass transfer while avoiding droplets so fine that they will not coalesce quickly. The dispersion then flows over a baffle into the settler, where a horizontal dispersion band deepens, droplets coalesce, and the phases separate. Heavy phase (usually aqueous) leaves under a draw-off baffle; light phase (usually organic) leaves over a weir. To understand why the aqueous-to-organic flow ratio and phase continuity matter so much here, see our resource on the O/A ratio in solvent extraction.

A single mixer-settler performs one equilibrium contact. Real circuits stack several in countercurrent cascade: organic flows from stage to stage in one direction while aqueous flows the opposite way, so fresh organic meets the leanest aqueous and loaded organic meets the richest. This countercurrent arrangement is what allows two or three physical stages to achieve recoveries that would need many more co-current contacts. For equipment-level detail on tank internals, baffles, and weir arrangements, see our mixer-settler equipment page.

What Are the Key Mixer Settler Design Parameters?

Mixer-settler design can be reduced to a handful of coupled parameters, each of which drives the others. The table below lists the values most commonly used in copper SX practice; all figures are typical industry values and must be confirmed with bench-scale or pilot testing for your specific liquor and reagent system.

Three of these deserve deeper discussion.

Mixer residence time. Two to three minutes per mix compartment is the entrenched industry norm for copper oxime extraction, and it reflects the kinetics of oxime extractants such as the LIX 984N and Acorga M5640 classes in aliphatic diluents. Fast-kinetics systems can sometimes justify shorter times, while viscous or modified reagents may need more. The mixer volume follows directly: V = Q_total × t, so a circuit treating 500 m³/h of combined phases at 2.5 minutes needs roughly 21 m³ of mixed volume per stage, before accounting for headspace and hold-up.

Specific settler area. Settler sizing is fundamentally an empirical exercise in phase disengagement. Designers size the settler plan area from the total phase flow divided by an allowable flow per unit area, and the allowable value depends on how quickly the particular organic/aqueous pair breaks. For conventional copper SX with clean liquors, dispersion band tests in the laboratory establish the capacity curve; the design point is then chosen so the band sits at a stable depth — commonly in the 150–350 mm range — with margin above flooding. If your liquor carries suspended solids or degradation products, expect the allowable loading to fall and the settler to grow.

Impeller power intensity and tip speed. These two numbers together define the shear environment. Enough power is needed to disperse the phases and keep solids suspended, but excessive tip speed creates fine secondary droplets that report to the settler as entrainment and, over time, stabilize crud. In copper SX mixer settler design, tip speeds are typically kept at or below roughly 3.5 m/s, and axial-flow pump-mix impellers are favored over high-shear radial designs. This is one reason some flowsheets move difficult separations to a centrifugal extractor, which achieves rapid phase separation mechanically instead of gravitationally.

How Many Stages Does Your Flowsheet Need?

Stage count is set by the extraction equilibrium, not by the equipment. The standard method is graphical: plot the extraction isotherm (equilibrium copper loading on the organic versus copper concentration in the aqueous) from shake-out tests, overlay the operating line defined by the O/A ratio and feed grades, and step off theoretical stages using the McCabe-Thiele construction. Our step-by-step walkthrough of the McCabe-Thiele method for solvent extraction shows exactly how this is done.

D2EHPA vs Cyanex 272: selection rules

Once the number of theoretical stages is known, divide by the expected stage efficiency to get the number of physical stages. Because real mixers never quite reach equilibrium, mixer settler design must include this correction. If the isotherm says 1.6 theoretical extraction stages are needed and the stage efficiency SX value for your system is 90%, the design needs two physical extraction stages.

For a conventional copper SX–EW circuit treating PLS from heap or agitated leaching, the resulting flowsheet is remarkably consistent across the industry:

• Extraction: 2 stages in series (occasionally 3 for low-grade or high-acid liquors), organic flowing countercurrent to aqueous.
• Stripping: 1 or 2 stages, where loaded organic transfers copper to spent electrolyte returning from electrowinning.
• Wash (optional): 1 stage to remove entrained aqueous impurities from loaded organic before stripping.

This “2 extract + 1–2 strip” pattern is typical, not universal. Nickel laterate circuits using D2EHPA style chemistry (and its analogs) or versatic acid derivatives, zinc circuits, and rare-earth separations often need more stages because their separation factors are lower. The general rule: the closer the operating line sits to the isotherm, the more stages you need, and the more sensitive recovery becomes to stage efficiency — which is exactly why our dedicated page on mixer-settler stage efficiency belongs in every stage-count study.

Which Mixer Settler Design Selection Rules Should You Follow?

Every project eventually reaches the point where the flowsheet must be committed to a configuration. The following decision rules express how experienced SX engineers make those calls; the phrase “preferred” here means lower lifecycle risk, not merely lower installed cost.

Mixer settler design decision rules:

1. For clean, high-flow copper PLS duty (above ~100 m³/h aqueous), a conventional box-type gravity mixer-settler is preferred because it offers the lowest cost per unit area, robust solids tolerance, and decades of operating precedent.
2. For low-flow, high-value or difficult-disengaging systems (below ~20 m³/h, or slow phase separation), a centrifugal extractor is preferred because its mechanical separation collapses residence time and slashes organic inventory.
3. For feeds with high suspended solids (above roughly 50–100 mg/L reaching SX), a mixer-settler with a dedicated crud-handling provision is preferred because centrifuges foul rapidly, whereas gravity settlers allow scheduled crud removal.
4. For corrosive copper electrowinning service, FRP (fiberglass-reinforced plastic) with an appropriate liner, or 316L stainless steel with acid-brick lining in hot acid zones, is preferred because carbon steel alone cannot survive hot sulfuric acid with ferric and chloride traces.
5. For circuits where organic losses dominate the operating cost, a larger settler area at lower loading is preferred because entrainment losses fall roughly exponentially with dispersion band residence and can exceed the cost of extra concrete and liner.
6. For campaigns with frequent duty changes or multiple feed sources, a modular mixer-settler design with adjustable weirs and baffle positions is preferred because hydraulic flexibility preserves performance as throughput and O/A ratios move.

The comparison below summarizes the main contacting options.

Why Do Entrainment and Phase Disengagement Shape the Whole Design?

Entrainment — organic droplets carried in the aqueous raffinate, or aqueous carried in the loaded organic — is the single largest hidden cost in a poorly sized circuit. Organic lost to the raffinate is purchased chemical leaving the plant; aqueous entrained into electrolyte carries chloride, silica, and iron that degrade cathode quality. Both failure modes trace back to phase disengagement behavior in the settler, which is why mixer settler design must treat the settler as a coalescence device first and a holding vessel second.

The physics is straightforward but unforgiving. Droplet rise velocity follows Stokes-type behavior, so halving droplet diameter roughly quarters its separation velocity. A mixer that shears organic into 200 µm droplets instead of 600 µm droplets demands roughly nine times the settling area for the same carryover. This is the practical meaning of the impeller tip-speed limit and of the recommendation to separate pump-mix and secondary mixing duties where kinetics allow. Crud — the sticky solids-plus-organic-and-fine-droplets interfacial layer that forms from colloidal silica, clays, and reagent degradation products — further slows disengagement and narrows the stable operating window. Crud management through skimming, hot washing, or dedicated treatment is a design input, not an afterthought; see our crud treatment in SX resource for the control toolbox.

Design responses to entrainment risk include: oversizing the settler relative to the clean-liquor test curve; specifying picket-fence baffles at the settler inlet to distribute the dispersion band evenly; installing organic skim and recovery launders on raffinate outlets; and returning recovered interfacial material to a treatment loop rather than to the mixer. A Zambian copper producer case (anonymized, illustrative figures) shows the leverage: by widening the settler by 20% and reducing tip speed from 4.1 to 3.2 m/s, a circuit cut raffinate organic carryover from an estimated ~120 ppm to below 50 ppm, recovering organic worth several hundred thousand US dollars per year at typical reagent prices. Treat those figures as directional — every circuit’s economics differ — but the mechanism is universal.

Why Is My Mixer-Settler Underperforming? Troubleshooting Failure Modes

When recovery drops, entrainment rises, or the interface wanders, the root cause is usually one of a small set of recognizable failure modes. Work through them in order:

1. Poor stage efficiency. Copper tenor in raffinate is higher than design, and the McCabe-Thiele step-off no longer closes → mixing energy too low, reagent concentration drift, or residence time eroded by increased flow → verify impeller speed and VFD settings against design, check organic composition by analysis, and re-run shake-out tests on current reagent batches.
2. Organic in raffinate (aqueous-side entrainment). Visible sheen or >50 ppm organics in the raffinate pond → excessive tip speed creating fine droplets, shallow dispersion band, or a flooded settler → reduce impeller speed toward ≤3.5 m/s tip speed, verify settler loading against the dispersion test curve, and inspect inlet baffling for short-circuiting.
3. Aqueous in organic (loaded organic contamination). Chloride or iron rising in tankhouse electrolyte, hazy loaded organic → over-pumping the dispersion into the organic weir, interface set too low, or crud stabilizing the band → adjust the interface level via the aqueous weir, increase wash-stage effectiveness, and schedule crud removal.
4. Crud accumulation. A thick, stable interfacial layer 100–300 mm deep that will not break → colloidal solids, clay fines, or reagent degradation products accumulating at the interface → install or restore skimming, route crud to hot-wash or centrifugal treatment, and tighten upstream solution clarification.
5. Phase reversal or “gunking”. The mixer suddenly runs aqueous-continuous when designed organic-continuous (or vice versa), with violent agitation and poor control → O/A ratio drift, feed flow swings, or impeller selection mismatched to phase continuity → correct the O/A ratio control loop, dampen feed surges, and confirm the impeller suits the intended continuous phase.
6. Weir and baffle level instability. Interface hunting between stages with periodic cross-contamination events → incorrect weir geometry for actual densities, or surging from pump-mix interaction → recalculate inter-stage weirs using measured densities at operating temperature and add level damping or control.
7. Accelerated liner or coating failure. Blistering, cracking, or liner loss within the first years → chemical incompatibility with hot acid or solvent attack on an unsuitable coating system → review mixer settler materials of construction against actual operating temperature, acid strength, and chloride levels, and upgrade to a tested FRP/lining system.

What Does a Mixer-Settler Cost? CAPEX Drivers and Ranges

Mixer-settler design decisions drive capital cost through three channels: tank volume (mixer plus settler plan area and depth), materials of construction, and internals complexity. Published cost data is scarce and site-specific, so the ranges below are illustrative for a conventional copper SX train and should be treated as order-of-magnitude only, with vendor quotations for any real project.

Mixer-Settler Settler Area and Mixer Volume Calculator, step by step

Two rules of thumb follow from the table. First, because settler area dominates, a mixer settler design that trades a modest mixer volume increase for a smaller, cleaner-settling dispersion (through gentler impeller duty) is often cash-positive. Second, materials escalation is nonlinear: moving from FRP to 316L can double the shell cost, and brick-lined steel sits between, so the corrosion assessment should be settled early with real liquor analyses rather than defaults.

Worked Mixer Settler Design Sizing Example

The following example walks through a complete first-pass sizing. All numbers are illustrative typical industry values, not guarantees; a real project must confirm them with bench and pilot testing.

1. Define the duty. Treat 120 m³/h (about 530 US gpm) of copper PLS at 2.0 g/L Cu with 1:1 O/A ratio in extraction. Total phase flow Q = 120 + 120 = 240 m³/h.
2. Set mixer residence time. Assume t = 2.5 min (typical for copper oxime systems). Required mixed volume per extraction stage: V = Q × t = 240 m³/h × (2.5/60) h = 10 m³ of dispersion. With one mix compartment per stage, specify ~12 m³ nominal mixer volume to allow headspace and start-up margin.
3. Check impeller sizing. Target power intensity 1.2 kW/m³ (illustrative) gives ~14 kW installed per mixer; for a 2.6 m mixer diameter, tip speed at, say, 90 rpm with a 1.2 m impeller is N·π·D = (90/60) × 3.14 × 1.2 ≈ 5.7 m/s — too high. Adjust to a larger-diameter, lower-speed impeller (~65 rpm, 1.4 m) giving ≈ 4.8 m/s, and refine with vendor data to bring the design below ~3.5 m/s at duty flow. This iteration is exactly why tip speed must be checked, not assumed.
4. Size the settler. Assume allowable loading 4.5 m³/h of total flow per m² (verified later by dispersion band testing). Settler plan area A = Q_total / loading = 240 / 4.5 ≈ 53 m² per stage. Choose a 12 m × 4.5 m settler (54 m²), with a 1.0 m total liquid depth giving roughly 3.4 min of settler residence on total flow.
5. Confirm dispersion band depth. At the design loading, assume the dispersion band stabilizes near 250 mm — comfortably inside the 150–350 mm typical window. If bench tests show the band exceeding ~300 mm at 4.5 m³/h·m⁻², either increase settler area (to ~5 m³/h·m⁻² or lower loading) or improve upstream mixing gentleness; design margin here is cheap insurance.
6. Close the stage count and materials. Suppose the isotherm and McCabe-Thiele construction require 1.8 theoretical stages; at 90% stage efficiency, specify 2 physical extraction stages plus 1 strip stage (2-strip if acid balance demands it). For duty at 30–40 °C with pH ~1.8–2.0 PLS, specify FRP shells with a vinyl-ester liner system, or 316L in organic-contact service with acid-brick-lined zones for hot spent electrolyte — final selection against the actual liquor analysis.

Six lines of arithmetic later, you have mixer volume, impeller constraint, settler area, dispersion margin, stage count, and a materials basis — the complete skeleton of a mixer settler design that can go straight into vendor enquiry.

Frequently Asked Questions

What is the typical residence time in a mixer-settler?
Each mixing compartment is normally sized for 2–3 minutes of residence time at total phase flow in copper solvent extraction, with 2.5 minutes a common design midpoint. Slow-kinetics systems or viscous organics may need 3–4 minutes, while fast oxime systems at elevated temperature occasionally justify less. The settler adds its own residence, often 15–40 minutes on total flow depending on area loading and depth. Residence time is set by kinetics, so confirm the value with bottle-roll and bench dispersion tests before freezing mixer volume.

How many stages does a copper SX mixer-settler plant need?
A conventional copper SX–EW circuit uses 2 extraction stages plus 1 or 2 strip stages, occasionally adding a wash stage after extraction. The exact count comes from a McCabe-Thiele construction on the measured extraction isotherm, divided by stage efficiency — typically 85–95% for copper oxime systems. Low-grade, high-acid, or impurity-heavy liquors sometimes justify a third extraction stage. Because each physical stage is a full tank, settling on one stage too many early in design is far cheaper than retrofitting later.

What is specific settler area and why does it matter?
Specific settler area is the total phase flow a settler can handle per unit of plan area, expressed in m³/h per m² (or equivalent). Typical copper SX designs work between about 3 and 8 m³/h·m⁻², but the allowable value is set empirically by dispersion band testing of your actual organic, aqueous, and crud load. Understating the area risks flooding, entrainment of organic in raffinate, and lost reagent; overstating it inflates capital cost. It is the single most site-specific number in mixer settler design.

How is stage efficiency defined in solvent extraction?
Stage efficiency expresses how close a real stage comes to equilibrium, usually as the ratio of actual concentration change to the change that perfect equilibrium would give, commonly 85–95% per mixer-settler stage in copper duty. It converts theoretical stages from an isotherm calculation into physical stages to build. Efficiency falls with poor mixing, reagent degradation, short residence, or severe entrainment. Track it by comparing raffinate grades against the McCabe-Thiele prediction, and investigate any sustained drop of more than a few percentage points.

Which materials of construction suit hot acidic SX service?
Copper SX mixers and settlers are most often built from FRP with a suitable liner, from 316L stainless steel, or from lined carbon steel with acid-brick protection in the hottest, most acidic zones. Selection depends on operating temperature, free acid, ferric and chloride levels, and whether the vessel sees organic, aqueous, or both. Hot spent electrolyte service is the most aggressive and usually justifies brick lining. Always qualify the material against your actual liquor analysis and maximum credible temperature, not against nameplate averages.

What causes crud and how should the design handle it?
Crud is a stable interfacial gunk formed when colloidal silica, clay fines, iron precipitates, and reagent degradation products accumulate at the aqueous-organic interface, typically growing to a layer 50–300 mm deep. It slows phase disengagement, stabilizes entrainment, and consumes organic. Design responses include upstream solution clarification, provision for interfacial skimming, dedicated crud treatment such as hot washing or centrifuging, and gentle impeller duty that avoids emulsifying crud back into the phases. Treat crud capacity as a required design feature in any solids-bearing feed.

When is a centrifugal extractor better than a mixer-settler?
For small flows (commonly below ~20 m³/h), slow-settling phase pairs, or high-value streams where low organic inventory and fast dynamics matter, a centrifugal extractor is preferred because it separates phases mechanically in seconds rather than by gravity over tens of minutes. It also suits confined spaces and intermittent operation. Its drawbacks are higher capital and maintenance cost, rotating-seal complexity, and poor tolerance of high suspended solids, so gravity mixer-settlers remain the default for large, dirty, cost-sensitive copper circuits.

Conclusion

Mixer settler design is ultimately a chain of a few disciplined choices: 2–3 minutes of mixer residence time, a settler sized from measured disengagement data, stage count set by the McCabe-Thiele method rather than by habit, impeller duty gentle enough to protect the dispersion band, and materials matched to the real liquor chemistry. Get those five decisions right and the circuit will deliver its design recovery for decades. If you are sizing a new train or debottlenecking an existing one, our team can help you de-risk the numbers — talk to us about pilot testing your liquor and reagent system before you commit steel. Browse more guides in our solvent extraction resource library.

References

1. U.S. Geological Survey, “Mineral Commodity Summaries — Copper”, USGS, 2025.
2. W.G. Davenport, M. King, M. Schlesinger, and A.K. Biswas, “Extractive Metallurgy of Copper”, 5th edition, Elsevier, 2011.
3. G.A. Kordosky, “Copper recovery using leach/solvent extraction/electrowinning technology: forty years of innovation, 22.5 million tonnes of copper annually”, JOM, 2002.
4. K.C. Sole, A.M. Sastre, et al., review chapters on solvent extraction in hydrometallurgy, in “Solvent Extraction and Liquid Membranes: Fundamentals and Applications in New Materials”, CRC Press.
5. Cytec Industries / Solvay, “Mining Reagents and Solvent Extraction Technical Handbook” (current edition, publisher level).
6. M. Cox, “Solvent extraction in hydrometallurgy”, chapter in “Solvent Extraction Principles and Practice”, Marcel Dekker / CRC Press.
7. J. Rydberg, M. Cox, C. Musikas, and G.R. Choppin (eds.), “Solvent Extraction Principles and Practice”, 2nd edition, Marcel Dekker, 2004.

Changelog

2026-10-21 — First published.