Thickening Area Calculation: Thickener Sizing Example

Last updated: 2026-10-30

The thickening area calculation for a 120 t/h feed of 28% w/w solids thickened to 62% w/w closes at 0.123 m² of settling surface per tonne of dry solids per day — 355 m² on test data, 426 m² after a 1.2 design factor, and a standard 24 m diameter thickener running a 0.78 m/h rise rate. Every intermediate number is worked out below from one batch settling test; swap in your own data and rerun the sequence.

This page is the hands-on companion to our solid liquid separation pillar, which covers the equipment family, and to the thickener design calculation overview. Here we stay on paper: the settling physics, three sizing methods, one fully worked example, and the mechanical checks that keep the chosen machine out of trouble.

Key Takeaways

• The unit area method converts one batch settling test into m² per tonne of solids per day.
• The worked duty of 120 t/h dry solids needs 355 m² of test area before any scale-up factor.
• A 1.2 design factor lifts that to 426 m², which selects a standard 24 m diameter bowl.
• Coe and Clevenger, Talmage and Fitch, and the solids flux design method bracket the answer within about 10%.
• With flocculant, the bottleneck usually shifts from settling area to feedwell rise rate — check both.

What Is Unit Area and Why Does It Govern Thickener Sizing?

A thickener is a gravity sedimentation vessel that converts a dilute feed slurry into a clarified overflow and a dense, pumpable underflow. Its economics hinge on one question: how much horizontal surface does the slurry need to surrender its solids at the rate the flowsheet demands? The thickening area calculation exists to answer that question in square meters before a single plate of steel is cut.

Unit area is a sizing metric that expresses the settling surface required per tonne of dry solids per day, written as m²/(t/d). A unit area of 0.12 m²/(t/d) means each tonne per day of solids needs 0.12 m² of bowl surface; a 2,880 t/d duty therefore needs about 346–426 m² depending on the scale-up factor applied. Unit area is the exchange rate between test work and steel, which is why every thickener vendor asks for it first.

The physics behind the number comes in three regimes. In free settling, dilute particles descend at their individual terminal velocities, governed by Stokes’ law, without interfering with one another. Hindered settling is a regime that begins once particles are crowded enough that each one drags on its neighbours’ fall — the slurry descends as a visible interface, and the velocity of that interface is the zone settling velocity, a function of local concentration rather than of particle size alone. Kynch’s classic 1952 theory showed that the whole batch settling curve is fixed by that concentration–velocity relationship, which is what makes a single cylinder test so information-rich. Below the zone-settling interface, compression settling is a consolidation stage that proceeds as the self-weight of the bed squeezes water out of a compressible mat — slow, and governed by bed yield stress rather than by Stokes’ law; the rheology behind it is treated in our slurry rheology guide.

One more definition anchors the arithmetic that follows. The compression point is the moment in a batch test that the descending interface stops decelerating smoothly and enters bed consolidation — visually, where the height–time curve bends from a straight, steep line into a flattening curve. Talmage and Fitch built their whole sizing construction around locating that point.

Which Sizing Method Should You Trust: Coe and Clevenger, Talmage and Fitch, or Solids Flux?

Three methods dominate thickener sizing practice, and they are mathematically cousins rather than rivals. All three reduce to the same statement: the applied solids loading, W/A, must not exceed the flux capacity of the slowest layer between feed and underflow concentration.

Coe and Clevenger (1916). The oldest unit area method runs settling tests at several concentrations between feed C0 and underflow Cu, measures a zone settling velocity u(C) at each, and takes the worst layer:

UA (C and C) = 24 × (1/C − 1/Cu) / u(C), maximized over C between C0 and Cu

with C in t/m³ and u in m/h, giving m² per tonne per day. The full Coe and Clevenger procedure is honest but expensive: four to six cylinder tests per duty, each at a different dilution.

Talmage and Fitch (1955). Their insight was that one batch test at the feed concentration already contains the whole Coe and Clevenger answer, provided you read the curve correctly. Locate the compression point, draw the tangent to the curve there, and find where that tangent crosses the underflow height Hu = H0 · C0/Cu. The time to that crossing, t_u, closes the sizing directly:

UA (T and F) = t_u / (C0 · H0)

with t_u in days, C0 in t/m³ and H0 in m — again m²/(t/d). One cylinder, one graph, one number.

Solids flux (Yoshioka). The flux method restates the same balance as a design diagram. Settling flux at any concentration is G = u(C) · C; the underflow withdrawal adds bulk transport, so the total flux a layer must carry is:

G_L = u(C) · C / (1 − C/Cu)

and the required area is A = W / G_L at the limiting concentration. Solids flux design is the version vendors plot for high-rate and paste machines, and it is the natural language for discussing paste thickener design, where compression becomes the governing term.

The ranges above are indicative values compiled from vendor-practice and handbook sources, stated to be confirmed by test work on your own slurry — never a substitute for a cylinder and a stopwatch. For first-pass screening they will put a duty within one standard diameter; for the datasheet, run the batch test.

How Does the Thickening Area Calculation Convert a Batch Test Into a 24 m Thickener?

Here is the full worked example. The duty: 120 t/h of dry solids arriving at 28% w/w, to be thickened to 62% w/w underflow. This is a conventional unflocculated design exercise — the pure area mathematics — and the flocculated variant is treated separately below.

how Solid-Liquid Separation is decided

Step 1 — Fix the basis. Dry solids W = 120 t/h = 2,880 t/d. Feed 28% w/w, underflow 62% w/w. Assume specific gravity of the dry solids of 2.7, typical for silicate gangue — state your own mineral density before reusing these numbers. Every concentration below is on a weight-percent basis converted to t/m³; the slurry densities follow from the mixture rule, 1/ρ = w_s/ρ_s + w_w/ρ_w.

Step 2 — Convert concentrations and flows. Feed slurry density: 1/(0.28/2.7 + 0.72/1.0) = 1/0.8237 = 1.214 t/m³. Feed concentration C0 = 0.28 × 1,214 = 339.9 ≈ 0.340 t/m³. Underflow slurry density: 1/(0.62/2.7 + 0.38/1.0) = 1/0.6096 = 1.640 t/m³, so Cu = 0.62 × 1,640 = 1,016.9 ≈ 1.017 t/m³. Volumetric feed: 428.6/1.214 = 353 m³/h. Underflow slurry: 120/0.62 = 193.5 t/h, or 193.5/1.640 = 118 m³/h. Water to overflow: 308.6 − 73.5 = 235 m³/h. These four flow figures feed every check later.

Step 3 — Run and read the batch test. A 1 L graduated cylinder is charged at C0 with an initial interface height H0 = 320 mm, flocculant-free, and the interface is recorded:

The first 8 minutes are linear: u0 = (320 − 224)/8 = 12.0 mm/min = 0.72 m/h. The curve then bends — slope 10 mm/min by 12 minutes, 6 by 15, under 3 by 20 — marking the compression point near t = 10 min, H = 200 mm. The underflow height is Hu = H0 × C0/Cu = 320 × 0.340/1.017 = 107 mm, which the suspension itself only approaches slowly.

Step 4 — Build the Talmage and Fitch construction. Draw the tangent to the curve at the compression point (10 min, 200 mm); it falls at 10 mm per minute, so its height is H = 200 − 10 × (t − 10). Where does it cross Hu = 107 mm? 200 − 10 × (t_u − 10) = 107, so t_u − 10 = 93/10 = 9.3 and:

t_u = 19.3 min

Step 5 — Compute the unit area and cross-check. Talmage and Fitch: UA = t_u / (C0 · H0) = (19.3/1440) / (0.340 × 0.320) = 0.01340 / 0.1088 = 0.123 m²/(t/d). Required test area: A = 0.123 × 2,880 = 355 m². The Coe and Clevenger single-point check at feed concentration gives A = W × (1/C0 − 1/Cu)/u0 = 120 × (2.941 − 0.983)/0.72 = 120 × 1.958/0.72 = 326 m² — an equivalent unit area of 0.113 m²/(t/d). The solids flux reading agrees with it: G_L = 0.72 × 0.340/(1 − 0.334) = 0.368 t/(m²·h) = 8.8 t/(m²·d), and 2,880/8.8 = 326 m². The two independent methods bracket 326–355 m², about 9% apart; the Talmage and Fitch figure governs as the conservative one.

Step 6 — Apply the design factor. Required design area = 1.2 × 355 = 426 m². The choice and defense of that 1.2 is the subject of the next section — it is not a fudge, and it is not negotiable.

Step 7 — Select diameter and check mechanics. Diameter = √(4 × 426/π) = √542 = 23.3 m, which rounds up to the standard 24 m bowl, 452 m² installed. The mechanical checks that justify the selection follow in the rise-rate and torque sections below.

What Design Factor Should the Thickening Area Calculation Carry?

The lab cylinder answers a smaller, kinder question than the plant asks. Wall effects in a 100 mm cylinder slow settling relative to a free surface; the lab test is quiescent while a plant bowl sees feedwell turbulence, raking stresses that shear the floc network, and concentration transients the cylinder never experiences. The published scale-up literature and vendor practice converge on a factor of roughly 1.2 for flocculated conventional and high-rate duties and up to 1.25 where the slurry is compressible or the feed grade swings. Our own convention: 1.2 on a clean, replicated batch test, 1.25 when the test work is a single cylinder or the ore is variable — and if a vendor’s sales sheet applies 1.15, ask what it removed to get there.

Here is the judgment that only a commissioned plant teaches. A 1.15-factored datasheet looks rigorous because it cites the test work, but year one is where the factor is spent: the first underflow density ramp pushes local bed densities well past the lab’s Cu, the rake torque climbs toward 90–95% of rating, and the operator manages the machine with bed pressure instead of design margin. We have inherited enough of these machines to say it plainly: the two or three meters of diameter that a 1.2 versus 1.15 factor costs are the cheapest insurance in the flowsheet. Buy them.

Design factors do not excuse thin test work, though. If the batch test was run at one concentration with no replication, treat the computed unit area as a hypothesis, not a result, and repeat the test at three concentrations the way the full Coe and Clevenger method intends.

How Do You Check Rise Rate and Rake Torque Before You Sign the Datasheet?

Area alone does not finish the thickening area calculation, because two more constraints can override it.

How to Select and Size a High Rate Thickener: selection rules

Rise rate. The pool must stay hydraulically quiet enough for fines to settle, which plant practice caps at roughly 0.5–1.2 m/h for mineral duty measured on total feed flow. The selected 24 m bowl receives 353 m³/h over 452 m², a rise rate of 0.78 m/h — comfortably inside the band. The 18 m alternative (254 m²) would run 1.39 m/h and blow the cap; the 20 m alternative (314 m²) would run 1.12 m/h with no rise-rate margin, and in any case falls short of the 426 m² the design factor demands. Only the 24 m bowl clears every constraint at once.

Rake torque. The mechanism must turn in a bed that is denser, on average, than the lab’s underflow figure suggests, because raking consolidates locally past Cu. As a first-pass screening figure, high-rate thickeners in base-metal tailings duty are commonly quoted between roughly 25 and 40 kNm per meter of bowl diameter — an indicative vendor-practice band, to be confirmed against the vendor’s own torque model. For the 24 m selection that is 600–960 kNm of duty, and the specification should carry a drive rated near 1,000 kNm. Two checks belong on the datasheet: rated torque at the normal operating bed level, and a lifting (post-stall) capacity of about twice the running figure. Underflow pumping limits and thickener rake torque both trace back to bed slurry rheology, so a rheology test on lab underflow at 62–66% w/w closes the mechanical case. For duties pushing past 65% w/w, the comparison in thickener vs filter press and the dewatering options in filter press for mining and decanter centrifuge for mineral slurry become relevant.

Where flocculant moves the answer. Dosing 15–25 g/t of polyacrylamide flocculant (indicative dose, to be confirmed by jar testing — see our flocculant dosage guide) typically lifts the zone settling velocity of a fine slurry from under 1 m/h to 3–5 m/h. Rerun the arithmetic with u0 = 4.0 m/h and the Coe and Clevenger area collapses to 120 × 1.958/4.0 = 59 m². The thickener does not shrink to 9 m, though, because the bottleneck moves: keeping 353 m³/h of feed under the 1.2 m/h rise-rate cap requires 294 m² — a 20 m flocculated machine. Flocculated sizing is therefore rise-rate-and-clarity governed, and the batch test on flocculated pulp must be read for overflow quality, not just interface speed. Where that flocculated underflow feeds a washing train, the area interacts with circuit wash efficiency as set out in our CCD wash ratio guide.

Eight Sizing Mistakes That Surface One Year After Commissioning

Each item below follows the same path: the symptom in the plant, the cause in the calculation, the fix in the method.

1. Full-scale capacity lands far below the lab unit area. Cause: the cylinder test ignores feedwell turbulence, raking shear and wall effects, and no design factor was applied. Fix: scale every computed area by 1.2–1.25 and replicate the batch test before freezing the diameter.
2. Overflow clarifies poorly from day one. Cause: area was settled on settling flux alone while the rise-rate cap was ignored. Fix: check Q_f/A against the 0.5–1.2 m/h band and re-select diameter if either constraint binds.
3. The rake stalls during the first density ramp. Cause: torque was estimated from the average underflow density, not from the consolidated bed the rake actually turns through. Fix: size the drive on the indicative 25–40 kNm/m band, confirm with the vendor’s torque model, and specify lift capacity.
4. Percent solids were used directly in the formulas. Cause: w/w figures were fed into equations that need volumetric concentration. Fix: always convert through slurry density first — C0 = 0.340 t/m³, not “28%”.
5. Underflow density was assumed, never verified. Cause: 62% w/w was quoted from a flowsheet precedent while the slurry is compressible and stops raking freely at 58%. Fix: run a lab consolidation or yield-stress test on the target underflow before committing the pump and drive.
6. Flocculant was added after the area was frozen. Cause: the duty was sized unflocculated, then a reagent scheme arrived and invalidated the whole basis. Fix: decide the reagent regime before sizing, since it changes both the governing method and the bottleneck.
7. The design factor was traded away to win the bid. Cause: a 1.15 or smaller factor made one standard diameter smaller and the price lower. Fix: treat factor erosion as a technical exception to be signed by the process owner, not a commercial line item.
8. Feed dilution was ignored. Cause: the feedwell was assumed to operate at C0 while the feed actually arrives denser, pushing the limiting layer up the column. Fix: verify feedwell dilution in the vendor’s design and keep the batch test concentration matched to the diluted condition.

Thickener Sizing FAQ: Questions Process Engineers Ask

What exactly does the thickening area calculation produce?

A required settling area in square meters, and usually its normalized form, the unit area in m² per tonne of dry solids per day. In the worked example the batch test gives 0.123 m²/(t/d), which against a 2,880 t/d duty means 355 m² of test area and 426 m² after the 1.2 design factor. That area, rounded up through standard diameters, is what selects the 24 m bowl.

How do I convert percent solids by weight into the units the formulas need?

Through the slurry density. With solids specific gravity 2.7, a 28% w/w slurry has density 1/(0.28/2.7 + 0.72/1.0) = 1.214 t/m³, so C0 = 0.28 × 1,214 ≈ 0.340 t/m³. Do the same at the underflow: 62% w/w gives 1.640 t/m³ and Cu ≈ 1.017 t/m³. Skipping this conversion is the most common arithmetic error in vendor-check reviews.

Why does the Talmage and Fitch method usually give a larger area than Coe and Clevenger?

Talmage and Fitch reads the whole answer from a single test at the feed concentration, and its tangent construction is sensitive to where you judge the compression point to be; a compression point read late produces a longer t_u and a bigger area. In the worked example it gave 355 m² against 326 m² from the single-point Coe and Clevenger check — about 9% apart. Taking the larger figure is deliberate conservatism, not inconsistency.

What design factor should I apply between the lab unit area and full scale?

Use 1.2 for a clean, replicated batch test on a conventional or high-rate flocculated duty, and 1.25 where the test work is a single cylinder, the ore is variable, or the slurry is strongly compressible. The factor compensates for wall effects, feedwell turbulence and raking shear that the cylinder cannot reproduce. Anything below about 1.15 should be treated as a technical exception with a named owner.

Does adding flocculant invalidate the batch test data?

It changes the design basis rather than invalidating it. Flocculant lifts zone settling velocity several-fold, so the settling-flux area collapses — but the thickener then usually becomes rise-rate limited, and the sizing question shifts to clarification and overflow clarity. Size the duty in the reagent regime the plant will actually run; an unflocculated batch test is useless for a flocculated machine and vice versa.

What rise rate should a properly sized thickener show?

Plant practice for mineral duty caps pool rise rate at roughly 0.5–1.2 m/h measured on total feed flow, with the low end where overflow clarity is critical and the high end where coarse, fast-settling solids dominate. The worked 24 m selection runs 353 m³/h over 452 m², or 0.78 m/h. If your computed rise rate sits above the cap, the diameter is set by hydraulics, not by the settling flux.

How is rake torque estimated before vendor datasheets exist?

As a screening figure, use an indicative 25–40 kNm per meter of bowl diameter for flocculated base-metal tailings duty — 600–960 kNm on the worked 24 m machine — then require the selected vendor to confirm with their own bed-model calculation. Specify the rated running torque at normal bed level plus a lifting capacity near twice the running figure, and have lab underflow rheology on hand to justify the number.

Conclusion: Closing the Thickening Area Calculation

The whole discipline of thickener sizing reduces to one cylinder and three closures. The batch test closes the settling physics; the unit area method — read through Talmage and Fitch, cross-checked with Coe and Clevenger and the flux construction — closes the area at 355 m²; the design factor and the rise-rate and torque checks close the equipment case at a 24 m, 1,000 kNm-class machine running 0.78 m/h. Run those steps on your own slurry and the thickening area calculation becomes arithmetic, not argument. The resource library carries the supporting guides for each step of the flowsheet, and if you want your batch data independently checked or turned into a sized proposal, request a quote with your test work attached and our process team will review the numbers before sizing anything.

Filter Press Area Calculator from Leaf Test Data

References

1. Coe, H.S. and Clevenger, G.H. (1916), “Methods for determining the capacities of slime-settling tanks”, Transactions of the American Institute of Mining Engineers, vol. 60.
2. Talmage, W.P. and Fitch, E.B. (1955), “Determining thickener unit areas”, Industrial & Engineering Chemistry, vol. 47.
3. Kynch, G.J. (1952), “A theory of sedimentation”, Transactions of the Faraday Society, vol. 48.
4. Yoshioka, N., Hotta, Y., Tanaka, S., Naito, S. and Tsugami, S. (1957), “Continuous thickening of homogeneous flocculated slurries”, Chemical Engineering (Japan), vol. 21.
5. Concha, F. (2014), Solid-Liquid Separation in the Mining Industry, Springer — thickening theory, batch test scale-up and flux theory.
6. Perry, R.H. and Green, D.W. (eds.), Perry’s Chemical Engineers’ Handbook, 9th ed., McGraw-Hill — liquid–solid operations and equipment, sedimentation.
7. Wills, B.A. and Finch, J.A., Wills’ Mineral Processing Technology, 8th ed., Butterworth-Heinemann — thickening and solid-liquid separation practice.
8. SME Mineral Processing Handbook, Society for Mining, Metallurgy & Exploration — sedimentation equipment selection and sizing practice.

Changelog

• 2026-10-30: First published.