Paste Thickener Design: Deep Cone Sizing & Underflow

Last updated: 2026-10-27

Paste thickener design is settled on bed depth and underflow yield stress, not on settling area alone: a deep cone vessel carrying 6–10 m of compression bed routinely holds underflow at 62–70 wt% solids where a high-rate thickener of the same diameter plateaus near 56–58%. This guide explains why, and how to size the vessel, rake and underflow pumps accordingly.

For the wider equipment landscape, start with our pillar page on solid-liquid separation; the area-based sizing mathematics shared by all gravity thickeners is covered in our thickener design calculation guide, and this article builds on rather than repeats it.

Key Takeaways

• Bed depth, not diameter alone, sets underflow density; self-weight compression does the final dewatering.
• Specify rake torque at the yield stress of the densest credible bed, not at design density alone.
• Cone angles shallower than roughly 25–30° compact poorly and leave the rake dragging paste uphill.
• Over-flocculated beds settle fast but compact badly; excess polymer raises bed strength and stalls pumps.
• Size underflow pumps on the measured flow curve (rheogram), never on water data or wt% alone.

What Is Paste Thickener Design?

Three definitions keep the vocabulary straight, because vendors use the same words for different vessels. A paste thickener is a deep-cone gravity settler that accumulates a deep compression bed and discharges a non-segregating underflow strong enough to support its own weight. A deep cone thickener is a steep-bottomed sedimentation vessel that uses a tall cylindrical shell above a conical floor, so the bed occupies most of the working volume and consolidates under self-weight. A high density thickener is an intermediate vessel that retains shallow-cone geometry but adds bed depth and a better flocculation system, typically reaching 55–62 wt%.

The process boundary between these machines is best drawn with rheology, not with wt% alone. Underflow yield stress is a material property that measures the shear stress the paste must exceed before it flows, and it rises steeply — often with the fourth to sixth power of solids concentration — as the floc network densifies. Industry usage treats underflows below roughly 10–15 Pa as conventional thickened tailings, 20–50 Pa as high-density discharge, and 40–100 Pa and above as true paste that does not re-segregate on a stack. These are typical ranges, to be confirmed by testwork; our slurry rheology guide explains the measurements.

A competent paste thickener design therefore fixes five things in sequence: the underflow solids target and its yield-stress window, the geometry that makes that density reachable (depth and cone angle), a flocculation regime that builds a compactable bed rather than a merely fast-settling one, a rake and drive rated for the worst bed the operators will ever create, and an underflow pumping and disposal interface sized on the rheogram. Each is treated below.

Key Design Parameters and Typical Ranges

The table below collects the parameters that govern the design, with typical ranges for metallurgical tailings duty. Every range is a starting point for testwork, not a substitute for it; mineralogy, particle size and water chemistry move all of them.

Two interactions in this table drive the economics in opposite directions. Flocculant dosage buys settling speed, which shrinks the diameter, but over-flocculated flocs are large and water-rich, so the dosage that saves steel can cost the last 3–5 wt% of underflow density and inflate the bed yield stress the rake must work against. Bed depth buys density cheaply (a taller shell costs far less than a wider one) but raises operating torque and the structural load on the drive cage. Good paste thickener design resolves both trades with compression and rheology testwork rather than rules of thumb.

Why Does Bed Depth — Not Area — Set Underflow Density?

This is where paste duty parts company with conventional thickener thinking, and the physics deserves spelling out. In the free-settling zone, concentration is set by whatever the feed delivers. But the underflow leaves from the bottom of the compression bed, where the solids have formed a continuous network and dewatering proceeds only under compressive stress. That stress comes from the weight of the bed above, roughly the buoyant density of the solids times the bed depth. The underflow concentration is therefore whatever point on the slurry’s compressive yield curve corresponds to the stress at the bottom of the bed — double the bed depth and you approximately double the consolidation stress, and the discharge density moves accordingly. Area controls throughput and overflow clarity; depth controls how dense the underflow can ever get.

practical guide: Thickening Area Calculation

Filter Press Area Calculator from Leaf Test Data

The practical consequence: underflow density shortfalls are almost never solved by re-running the unit area calculation. If a vessel plateaus at 60 wt% against a 65 wt% target, the honest options are deeper bed, longer residence, changed flocculation, or a different vessel.

Side-wall angle matters for the same reason. In a deep cone with a floor at 30–45° from horizontal, the compression bed extends down into the cone, and solids slide gently toward the discharge while still under full compressive load. Make the cone too shallow and three things go wrong at once: the effective compression bed retreats into the cylindrical shell, so the paste loses depth exactly where it should be gaining it; the rake must physically drag dense paste up a long flat floor, which multiplies torque; and the underflow draw-off sits where compaction is weakest. A deep cone with too shallow a cone angle compacts poorly, no matter how tall the shell above it — which is why vendors move to 45–60° floors where backfill densities of 70 wt% and above are required.

The same non-linearity governs the drive. Because bed yield stress scales steeply with solids concentration, a bed that drifts 2–3 wt% above the design underflow density can double the torque demand — and beds drift whenever the ore changes or operators chase density. The rake and drive must therefore be rated against the yield stress of that “stalled bed” scenario, not against the torque implied by design underflow density on an average day.

Paste Thickener Design Workflow: Six Steps From Feed Data to Drive Rating

The workflow below takes a process engineer from a mass balance to a defensible equipment specification. Steps 2, 3 and 5 are where paste thickener design differs most from ordinary sizing, and all three lean on measurements a conventional settling programme never collects.

1. Fix the mass balance and the underflow target. Establish dry solids feed rate, feed solids concentration, and the underflow specification the downstream process actually needs — expressed both in wt% and as an acceptable yield-stress window (for example 60–70 wt% and 30–80 Pa, “typical range — to be confirmed by testwork”). Write down the disposal interface too: stackable discharge, backfill, or filtration, because it sets how hard the rheology target is.

2. Run compression and rheology testwork, not just settling tests. The minimum programme: settling rate tests for flux sizing; compression-column tests giving equilibrium bed density versus depth; and flow curves (typically Herschel–Bulkley) at three or four solids concentrations spanning the candidate underflow range. Without the flow curves you cannot rate the rake, size the pumps or set the yield-stress window.

3. Choose the geometry: diameter from flux, depth from compression, angle from density. Size the area on a design solids flux of roughly 3–10 t/m²/d, then set bed depth from the compression curve so the bottom-of-bed stress delivers the target underflow density with margin. Select the cone angle so the bed stays in compression down to the discharge — 30–45° for most paste duty, steeper for backfill.

4. Design the feedwell and the flocculant regime for compactability. Feed dilution, gentle mixing energy and (on many feeds) staged polymer addition produce dense, tight flocs that settle adequately and compact well. Resist the temptation to dose polymer until the bench test looks fastest: the fastest-settling floc is usually the worst-compacting bed. An anionic polyacrylamide at 15–40 g/t is the usual starting envelope.

5. Rate the rake and drive at yield stress. Compute the running torque expected at design bed depth and density, then re-compute it for the stalled-bed scenario — the densest, highest bed the operators could plausibly create, usually 2–3 wt% above target underflow density. Rate the drive at 2.5–3.5× the running estimate, with a rake-lift cylinder rated to raise the full cage out of a compacted bed. The classical K-factor shortcuts are a sanity check at best; our thickener design calculation guide covers them, and they understate paste duty.

6. Engineer the underflow system and the commissioning checks together. Select pumping on the rheogram (next section), then lock in a commissioning checklist: bed-level instrument calibrated against a manual survey, torque baseline trended across the first bed build, floc quality audited against the bench reference, underflow line pressure logged from day one, and overflow turbidity trended through the first full-density week. Commissioning a paste thickener is a commissioning of the rheology model as much as the machine.

When Does Paste Thickening Beat High-Rate Thickening?

The selection logic is narrower than vendors sometimes suggest, and each scenario has a firm conclusion rather than a shrug. The table summarises the boundaries; the scenarios explain them. For the full trade against pressure filtration, see our thickener vs filter press comparison.

practical guide: How to Select and Size a High Rate Thickener

Scenario 1 — water recovery is the driver. Compare a 55 wt% high-rate underflow against 65 wt% paste on 10,000 t/d of dry solids: the paste vessel leaves roughly 5,380 t/d of water in the underflow versus 8,180 t/d, so about 2,800 m³/d of additional water reports to the process circuit. Where make-up water is scarce, that difference routinely justifies the deeper vessel alone; where water is abundant, a high-rate machine is the correct answer.

Scenario 2 — tailings are dry-stacked or discharged centrally at moderate scale. Below roughly 15,000 t/d of dry solids, a single paste thickener producing a non-segregating discharge is usually the cheapest stackable solution, and the deposit desiccates far faster than conventionally thickened slurry. Above that tonnage the vessel count multiplies and a high-density thickener feeding pressure filtration often wins; our dry-stack tailings page works through that boundary.

Scenario 3 — the underflow feeds backfill. Hydraulic and paste fill plants need 68–74 wt% with pumpable rheology, and only a deep cone geometry reaches it with acceptable flocculant consumption. This is the strongest pro-paste conclusion in the list: a high-rate thickener cannot deliver the duty.

Scenario 4 — the duty is washing or CCD. Compression is irrelevant to wash efficiency; what matters is wash ratio per stage, covered in our CCD wash ratio guide. Keep high-rate thickeners in CCD trains, and treat a paste proposal on a CCD circuit as a red flag from the vendor.

Scenario 5 — the feed is fine and clay-rich. Where a large slimes fraction dominates, paste duty demands extreme polymer dosages and still yields marginal densities; the economical route is a high density thickener at 55–60 wt% followed by filtration if a stackable product is required. Do not force paste duty on a feed whose compression curve refuses it.

What Can Go Wrong — and How Do You Fix It?

1. Underflow density plateaus below target. Symptom: underflow holds steady at 60–62 wt% against a 66 wt% design value. Cause: bed operated too shallow, loose fast-settling flocs, or a compression curve that was never measured so the target was optimistic. Action: raise the bed in 0.5–1.0 m steps while trending density and torque; re-trim dosage toward tighter flocs; if density still will not move, re-run the compression testwork — the fix may be a realistic target, not more steel.

2. Torque trend climbs and trips follow an ore change. Symptom: drive torque at normal bed level rises 50–100% over days after the mill starts treating a finer or higher-clay ore. Cause: the new ore builds a bed with higher yield stress at the same wt%. Action: raise rakes, dig the bed down through elevated underflow extraction, restore a shallower bed, and re-derive the drive margin against the new flow curve.

3. Over-flocculated bed — stringy floc and stalled underflow pumps. Symptom: large, watery flocs in the feedwell; underflow pumps lose suction pressure and stall even at normal bed level. Cause: polymer over-dose builds an elastic floc network that settles fast but resists both compaction and pipe flow. Action: step dosage down in 2–3 g/t increments while watching density, torque and pump pressure; re-audit the make-down plant, because poorly aged polymer drives over-dosing.

4. Underflow pump pressure oscillation and dry firing. Symptom: discharge pressure swings widely; paste surges then trickles. Cause: the pump was selected on water-duty charts or wt% alone, and the actual flow curve sits close to the pump’s stall point. Action: re-select the pump against the measured rheogram — positive-displacement or high-thrust designs for the upper paste range — and add a density-based interlock.

5. Coarse sanding and bed tunnelling in the cone. Symptom: hard, localised accretion on the cone floor; periodic torque spikes; density swings. Cause: coarse fraction short-circuits from the feedwell to the floor before flocculation, or a floor slope and rake profile that cannot transport the coarse tail. Action: review feedwell dilution, verify upstream classification, and confirm the cone angle and rake profile against the coarse fraction — sanding is a geometry problem far more often than a dosage problem.

6. Overflow turbidity spikes after commissioning. Symptom: clear overflow for days, then persistent cloudiness. Cause: bed level allowed to rise through the clarification zone during density chasing, or a blanket turnover after a feed surge. Action: enforce a hard bed-level high alarm with feed-rate cutback, and treat the clarification zone as a surge buffer, not a polishing pond.

What Does Paste Thickener Design Cost?

Costs below are indicative supply-side budgeting figures shaped by recent equipment packages and industry norms; every project needs a quotation, and civil works dominate site variance. Assumptions: throughput 3,000–15,000 t/d dry solids, tailings duty, FOB supply excluding duties, freight, foundations and electrical infrastructure, US dollars.

Capital: the equipment package (vessel, rake and drive, feedwell, flocculant plant, instrumentation) for an 18–25 m paste thickener typically falls in the range of US$2.5–5.0 million — roughly 1.5–2.5× the equivalent high-rate thickener, the premium being shell steel, drive torque and cone engineering. Foundations commonly add 30–60%. Positive-displacement or high-thrust underflow pumping for true paste duty adds roughly US$0.5–1.5 million on longer runs.

Operating: flocculant at 15–40 g/t and US$2.5–4.0/kg delivered costs US$0.04–0.16 per tonne of dry solids — the largest consumable and the item with the most optimisation leverage. Drive power for an 18–26 m vessel runs 90–250 kW; at US$0.10/kWh and 8,000 operating hours, roughly US$70,000–200,000 per year, with underflow pumping adding materially more than in a high-rate circuit because paste moves slower against higher pressure. Maintenance is dominated by the underflow pumps and the flocculant plant, not the rake.

Indicative Field Data (Anonymised)

The following figures are an illustrative composite drawn from anonymised project commissioning experience, not attributable to any named client. A 22 m deep cone thickener on gold tailings, 4,800 t/d dry solids, target underflow 65 wt%: raising the operating bed from 4.2 m to 7.5 m lifted underflow density from 61.5 wt% to 66.8 wt%, with yield stress moving from 28 Pa to 68 Pa — squarely into paste territory. Flocculant dosage was trimmed from 28 g/t to 21 g/t after the make-down plant was corrected, with no loss of overflow clarity. Running torque settled at 18–24% of drive rating, validating the 3× margin specified at yield stress.

Frequently Asked Questions

What underflow yield stress should a paste thickener target?

Set the target as a window, not a number: typically 30–80 Pa at the discharge for stackable paste duty, wide enough to absorb ore variation yet low enough for the pumps. A well-executed paste thickener design fixes this window in testwork, then holds the vessel inside it through bed-level and dosage control.

How deep should the compression bed be?

Deep enough that the self-weight stress at the bed bottom delivers the target underflow density on the slurry’s compression curve — usually 4–10 m on tailings paste duty. There is no universal depth: the compression testwork gives the answer, and the shell is bought tall enough with 1–2 m of margin.

Can an existing high-rate thickener be converted to paste duty?

Sometimes, for modest targets. Raising bed depth and upgrading the flocculant system can move a high-rate vessel from 55 to 60–62 wt%, but the shallow cone, drive rating and rake profile limit anything further. Conversions beyond that are rarely economic; a purpose-built deep cone vessel reaches 65–70 wt% far more reliably.

What should a paste thickener commissioning checklist include?

At minimum: bed-level instrument verified against manual survey; torque baseline recorded through the first full bed build; floc quality compared with the bench reference; underflow density, line pressure and pump performance trended together; and the flocculant make-down plant audited for ageing time and mixing energy. Each item validates one design assumption.

Why must underflow pumps be sized on the rheogram rather than on water?

Paste is a yield-stress fluid: in laminar pipe flow the pressure gradient includes a direct yield-stress term that water charts never capture, and a pump sized on water equivalent can be short by several hundred kilopascals at design flow. Centrifugal sets also lose efficiency sharply as yield stress approaches stall. Only the measured flow curve gives an honest selection.

What flocculant dosage does paste duty typically require?

Anionic polyacrylamide at roughly 15–40 g/t of dry solids, higher than high-rate thickening because the goal is a compactable bed rather than maximum settling velocity. The optimum shifts with feed concentration, water hardness and P80, and must be re-verified whenever ore changes. Over-dosing is the most common operating error on paste duty.

How does paste thickening change tailings water recovery?

Substantially, and that is usually the economic case. Moving underflow from 55 to 65 wt% on 10,000 t/d of dry solids returns roughly 2,800 m³/d more water to the process circuit — water that otherwise evaporates on a dam or must be pumped from a remote source. On water-scarce sites this alone can justify the deeper vessel.

Is a paste thickener always better than a filter press for dry stacking?

No. A paste thickener produces a stackable, non-segregating sludge; a filter press produces a handleable cake at 75–85 wt% solids. For sites needing vehicular traffic on the deposit, filtered cake is often mandatory, and the thickener serves as the press’s efficient first stage. The two are complements more often than competitors.

Conclusion

Good paste thickener design refuses to treat the vessel as a big settling tank. Bed depth — not area — sets what the underflow density can ever be; the cone angle decides whether the bed stays in compression down to the discharge; the drive earns its rating against the yield stress of the worst bed an operator can build; and the underflow pumps obey the rheogram, not the water chart. Get those four measured and specified honestly, and the vessel will deliver 62–70 wt% underflow for the life of the mine. If you are selecting a deep cone thickener for dry-stack, backfill or high water-recovery duty, our team can run the compression and rheology testwork that de-risks the specification — request a proposal through our pilot testing services.

References

1. Jewell, R.J. & Fourie, A.B. (2015). Paste and Thickened Tailings — A Guide, 3rd edition. Australian Centre for Geomechanics, University of Western Australia, Perth. acg.uwa.edu.au
2. Nguyen, Q.D. & Boger, D.V. (1998). “Application of rheology to solving tailings disposal problems.” International Journal of Mineral Processing, Vol. 54, pp. 217–233.
3. Boger, D.V. (2012). “Rheology of Slurries and Environmental Impacts in the Mining Industry.” Annual Review of Chemical and Biomolecular Engineering, Vol. 3, pp. 127–149.
4. Concha, F. (2014). Solid–Liquid Separation in the Mining Industry. Fluid Mechanics and Its Applications, Springer, Cham.
5. Wills, B.A. & Finch, J.A. (2016). Wills’ Mineral Processing Technology, 8th edition. Butterworth-Heinemann, Oxford.
6. Mular, A.L., Halbe, D.N. & Barratt, D.J. (eds.) (2002). Mineral Processing Plant Design, Practice, and Control. Society for Mining, Metallurgy and Exploration (SME), Littleton, Colorado.

Changelog

2026-10-27 — first published.