DAF vs Sedimentation: Choosing the Right Solid-Liquid Separator

What is the fundamental difference between DAF and sedimentation?

Both processes exploit the same physics — the terminal velocity of a particle moving through a fluid under a body force — but they apply it in opposite directions. In gravity sedimentation (a primary or secondary clarifier), the body force is gravity acting on particles denser than water, and the solids move downward to a collection hopper while clarified water rises to an overflow weir. In dissolved air flotation (DAF), micro-bubbles (10–100 µm) generated by depressurising an air-saturated recycle stream attach to and become enmeshed in the flocs, lowering their effective bulk density below that of water so the aggregates move upward to a surface float layer that is skimmed away.

The consequence is that the same particle can be a poor candidate for one process and an ideal candidate for the other. A dense mineral floc (specific gravity 1.05–1.20) settles readily but resists flotation; a light biological floc, an algal cell, or an oil-water emulsion droplet (specific gravity 0.85–1.01) barely settles but floats efficiently once seeded with bubbles. The engineering decision therefore begins not with the equipment catalogue but with the density and character of the solids you must remove. For background on the flotation mechanism itself, see our overview of what a DAF system is and how it works.

What does Stokes' law tell us about the choice?

The terminal (rise or settling) velocity of a discrete spherical particle in the laminar regime (particle Reynolds number Re_p < 1) is given by Stokes' law. It is the single most useful equation for framing the DAF-versus-sedimentation decision because its sign tells you the direction of separation and its magnitude tells you the surface loading rate you can afford.

Two features drive the design. First, velocity scales with the square of diameter, so coagulation and flocculation — which grow a 5 µm colloid into a 500 µm floc — change settling velocity by up to four orders of magnitude. This is why chemical conditioning is inseparable from clarifier design in both technologies; see our guide to coagulation and lamella clarifier design for the conditioning side.

Second, velocity scales linearly with the density difference (ρ_p − ρ_w). When this term is small and positive, gravity settling is slow and DAF wins; when it is negative (a buoyant particle), only flotation works; when it is large and positive, sedimentation is both faster and cheaper. DAF, in effect, engineers the density difference: a floc of ρ_p = 1010 kg/m³ that would settle at a feeble 0.03 mm/s can be turned into a bubble-floc aggregate of effective density 600–900 kg/m³ that rises at 5–15 mm/s — hundreds of times faster.

How do surface loading rate and overflow rate compare?

The sizing currency for every clarification process is the surface loading rate — flow divided by separation area — because, for ideal continuous separation, a particle is captured only if its terminal velocity exceeds the upward (or the net downward) fluid velocity. In a settling tank this threshold is the overflow rate (or surface overflow rate, SOR); in DAF it is the hydraulic loading or rise rate. The critical particle is the one whose v_t exactly equals the loading rate.

The practical ranges differ by nearly an order of magnitude, and this is the headline reason DAF footprints are so much smaller:

A conventional round clarifier loaded at 1.5 m/h needs roughly ten times the plan area of a DAF loaded at 12 m/h treating the same flow. That footprint advantage — driven entirely by the higher achievable bubble-floc rise velocity — is often the deciding factor on a congested industrial site.

Which solids favour flotation and which favour settling?

Match the process to the particle. The discriminating property is the effective specific gravity of the conditioned floc, followed by its tendency to entrain or generate gas.

Favour DAF (flotation):

• Fats, oils and grease (FOG) and oil-water emulsions — droplets are lighter than water (SG 0.85–0.95) and rise naturally; bubbles accelerate and stabilise the process. DAF is the default for oil and grease separation.
• Algae and cyanobacteria — near-neutral buoyancy and gas vacuoles make them notoriously slow to settle but easy to float; DAF is standard in algal-laden surface-water works.
• Low-density biological and hydroxide flocs — light activated-sludge floc, and gelatinous metal hydroxides from coagulation, float more reliably than they settle.
• Fibrous and paper-mill solids, and streams prone to rising sludge from denitrification gas.

Favour gravity sedimentation:

• Dense inorganic and mineral solids — grit, silt, sand, precipitated carbonates and metal salts (SG 1.2–2.6) settle fast and cheaply.
• High solids concentrations (>1,000–3,000 mg/L) where zone/hindered settling and thickening are needed — clarifiers double as thickeners and reach higher underflow solids.
• Very high flows where the energy cost of saturating a recycle stream becomes significant relative to a passive gravity basin.

A useful rule of thumb: if the conditioned floc has a settling velocity below ~1 m/h, or a specific gravity within ~0.02 of water, seriously evaluate DAF; if it settles above ~3 m/h, sedimentation is likely cheaper to build and run.

How does coagulant dose change floc density and the decision?

Coagulant chemistry does not merely aggregate colloids — it sets the density of the resulting floc, and therefore whether it settles or floats. Metal-salt coagulants (alum, ferric chloride, PACl) work partly by charge neutralisation and partly by sweep flocculation, in which colloids are enmeshed in bulky, highly hydrated amorphous hydroxide precipitates — Al(OH)₃ or Fe(OH)₃. These sweep flocs are large but extremely open, with bound water fractions above 95%, giving an effective specific gravity often only 1.001–1.02. Such flocs settle sluggishly but float beautifully, which is precisely why DAF is so often paired with ferric or alum coagulation on coloured, low-turbidity or algal waters.

Increasing the coagulant dose in the sweep-floc regime grows floc size (helping Stokes velocity via d²) but simultaneously lowers floc density (hurting the density-difference term). The two effects partly cancel for settling, but both help flotation, because a larger, lighter floc presents more area for bubble attachment and needs less buoyant lift per unit mass. Polymer (polyacrylamide) flocculant behaves differently: bridging flocculation produces denser, stronger, shear-resistant flocs that both settle faster and — because they resist break-up in the pressure-release nozzle — hold bubbles well. Our note on coagulation and flocculation for clarifiers covers dose optimisation in more detail. The practical lesson: run comparative jar tests and bench float tests, because the same coagulant regime can push a water toward either technology depending on dose and pH.

Worked example: comparing required area for a light floc

Consider a surface water dosed with ferric chloride, producing sweep flocs with a mean diameter d = 150 µm (1.5×10⁻⁴ m) and floc specific gravity 1.005 (ρ_p = 1003 kg/m³). The design flow is Q = 100 m³/h at 15°C (μ = 1.14×10⁻³ Pa·s, ρ_w = 999 kg/m³).

Step 1 — settling velocity by Stokes' law:
v_t = (9.81 × (1.5×10⁻⁴)² × (1003 − 999)) / (18 × 1.14×10⁻³)
= (9.81 × 2.25×10⁻⁸ × 4) / (2.05×10⁻²)
= 8.83×10⁻⁷ / 2.05×10⁻² = 4.3×10⁻⁵ m/s = 0.155 m/h.

Step 2 — area if we settle it: to capture this floc by gravity, SOR ≤ v_t = 0.155 m/h, so A = Q / SOR = 100 / 0.155 = 645 m². That is an impractically large basin (a 28.6 m diameter circular clarifier) for a modest 100 m³/h.

Step 3 — area if we float it: attach micro-bubbles so the aggregate reaches an effective specific gravity of 0.75 (ρ_eff = 749 kg/m³) at an effective aggregate diameter of 200 µm. The rise velocity magnitude is
v_r = (9.81 × (2.0×10⁻⁴)² × (999 − 749)) / (18 × 1.14×10⁻³)
= (9.81 × 4.0×10⁻⁸ × 250) / (2.05×10⁻²) = 9.81×10⁻⁵ / 2.05×10⁻² = 4.8×10⁻³ m/s ≈ 17 m/h.

In practice a DAF is designed at a conservative hydraulic loading of 10–12 m/h to allow for the flow of the recycle and non-ideal hydraulics, giving A = (Q + Q_r) / SLR. With 50% recycle, A = 150 / 12 = 12.5 m². The DAF requires roughly 50 times less plan area than the gravity basin for this light floc — a decisive result that follows directly from turning a +4 kg/m³ density difference into a −250 kg/m³ one. For the recycle and saturator side of that calculation, see our detailed DAF sizing walk-through.

Worked example: comparing performance for a dense floc

Now take a lime-softening or metal-precipitation sludge with dense flocs: d = 200 µm, floc specific gravity 1.10 (ρ_p = 1099 kg/m³), same water properties, same 100 m³/h.

Settling velocity: v_t = (9.81 × (2.0×10⁻⁴)² × (1099 − 999)) / (18 × 1.14×10⁻³) = (9.81 × 4.0×10⁻⁸ × 100) / 2.05×10⁻² = 1.9×10⁻³ m/s = 6.9 m/h.

Area if we settle it: designing a clarifier at SOR = 2.5 m/h (well below v_t, for margin and thickening) gives A = 100 / 2.5 = 40 m² — a 7.1 m diameter clarifier that will also thicken the sludge to 2–6% dry solids in its base. A lamella settler would shrink the footprint further to ~15–20 m² of floor space by stacking inclined plates; see our primary clarifier design guide and lamella clarifier design guide.

Here DAF offers no advantage: the dense floc settles at nearly 7 m/h, faster than a typical DAF rise rate, and gravity separation is passive (no saturator, no recycle pump, no compressed air). Trying to float an SG-1.10 floc would demand an unrealistically high air-to-solids ratio to overcome its excess mass. To lift one kilogram of this floc, the buoyant deficit that must be supplied by attached air is proportional to the excess density (100 kg/m³ here), which is two orders of magnitude larger than the buoyant deficit of the light floc in the previous example — so the air demand, and hence saturator energy, would be prohibitive. For this water, sedimentation is both technically superior and cheaper to own.

The two worked examples bracket the decision cleanly. In the first, a +4 kg/m³ density difference gave a hopeless 645 m² settling area but a compact 12.5 m² DAF; in the second, a +100 kg/m³ difference gave a practical 40 m² settler and no sensible flotation route. The crossover between the two regimes is not a matter of taste or supplier preference — it is fixed by the density difference term in Stokes' law, and it moves only when you change the chemistry that sets the floc density. This is why the very first design task is always to characterise and, where possible, deliberately engineer the conditioned floc, rather than to pick a technology and force the water to suit it.

How do footprint, start-up, sludge solids and cost compare?

Beyond the separation physics, four practical dimensions usually decide the final specification.

DAF's rapid start-up and surge tolerance make it well suited to batch and intermittently loaded industrial effluents (breweries, dairies, food factories) where a settling blanket would never stabilise. Sedimentation's low energy demand and thickening ability favour continuous, high-flow, high-solids municipal and mineral duties. For high-value or hydraulically awkward flotation duties, a compact packaged dissolved air flotation unit or a bespoke clarifier can be specified from the same supplier; discuss the duty with the water treatment equipment team before committing to a footprint.

When is a hybrid — ballasted settling or a DAF-clarifier — the best answer?

The DAF-versus-sedimentation choice is not always binary. Several hybrid processes engineer the density difference deliberately to capture the best of both.

• Ballasted sedimentation (e.g. microsand-ballasted flocculation) injects fine, dense microsand (SG ~2.6) that becomes embedded in the floc, raising ρ_p dramatically. Settling velocities climb tenfold, allowing SORs of 40–80 m/h — comparable to or faster than DAF, in a compact settler — with the sand recovered by hydrocyclone and recycled. Ideal for very high peak flows and stormwater.
• DAF-clarifiers / combined units take a float layer off the top and a settled fraction from the bottom of the same tank, useful where a stream carries both buoyant (FOG) and dense (grit, precipitate) solids simultaneously.
• Counter-current / high-rate DAF re-engineers the contact and separation zones to push hydraulic loading to 30–40 m/h, closing the footprint gap with ballasted settling for light-floc waters.
• Lamella-enhanced settling adds inclined plates to a conventional clarifier, increasing effective settling area 5–10× within the same tank volume for dense flocs — the cheapest route to a small settling footprint when the solids are heavy.

The selection logic is therefore layered: first establish floc density and character; then, if the floc is light, compare DAF against ballasted settling on footprint, energy and complexity; if the floc is dense, compare conventional against lamella settling on footprint and thickening. Only a bench programme — jar tests for chemistry, settling-column tests for v_t, and float tests for air-to-solids response — closes the decision with confidence. The equations tell you which region of the design space you are in; the bench tests tell you exactly where within it.

How to select between DAF and sedimentation

1. Characterise the solids. Measure or estimate conditioned floc size, specific gravity, and any buoyancy/gas tendency (FOG, algae, biofloc). This sets the sign and magnitude of Stokes velocity.
2. Estimate the terminal velocity. Apply Stokes' law to the conditioned floc. A settling velocity below ~1 m/h or a near-neutral density points to flotation; above ~3 m/h points to sedimentation.
3. Compare required areas. Compute A = Q / SOR for settling and A = (Q + Qr) / SLR for DAF using realistic loading rates. Compare footprints on the actual site constraint.
4. Weigh opex, start-up and sludge solids. Balance DAF energy and fast start-up against sedimentation's passive operation and thickening ability, and check the downstream dewatering duty.
5. Consider hybrids. For light flocs, evaluate ballasted settling and high-rate DAF; for dense flocs, evaluate lamella settling before defaulting to a large conventional basin.
6. Confirm by bench testing. Run jar tests, settling-column tests and bench float tests on the real water to fix chemistry, loading rate and air-to-solids ratio before procurement.