DAF Air-to-Solids Ratio: Sizing Air, Not Tank Volume

A dissolved air flotation unit rarely fails because its tank is too small. It fails because the air-to-solids ratio at the moment of the upset was below what the floc needed, and no amount of residence time will float a particle that never acquired enough attached air. The A/S ratio, kilograms of released air per kilogram of solids fed, is the number that decides whether a DAF meets consent or pushes solids into the sub-natant. The failure mode is expensive in a specific way: the tank looks correct on the general arrangement, the unit passes commissioning on a mild feed, and then the first strong load of the season collapses the float blanket and sends TSS and FOG straight onto the downstream biology.

What the air-to-solids ratio actually measures

Bubble-particle attachment is a surface phenomenon, so A/S is really a proxy for the bubble surface area and buoyant volume presented per unit mass of solids. Below the required ratio, a fraction of the flocs never reach positive buoyancy and report to the underflow regardless of how long they stay in the tank. Above it, the blanket is robust but you are paying for air, and for the recycle pumping that delivers it, that the process does not need.

Mass alone is not the point. Because at least one bubble is needed for every particle to be floated, how the air is divided matters more than how much of it there is. Water Corporation's DS211 design guideline makes the arithmetic vivid: a single 2 mm bubble holds the same air as 64,000 bubbles of 50 micron. The same released mass either floats the load or does almost nothing, depending on bubble size and uniformity.

How attachment happens also changes what enough air means. DS211, drawing on Haarhoff and van Vuuren, distinguishes three mechanisms: adhesion of bubbles to preformed flocs, which dominates when particles are similar in size to the bubbles; mechanical enmeshment, where particles are much larger than the bubbles and the bubbles lodge in the jagged floc edges; and entrapment of bubbles into flocs still growing, which needs high particle concentration and rapid flocculation. A design that assumes one mechanism and gets another will miss its air target even when the arithmetic checks out.

A DAF does not float solids. It floats bubble-floc agglomerates, and the air-to-solids ratio is the only design number that says whether enough of them will form.

The governing equation, term by term

The classic flotation design equation, in the Metcalf and Eddy / Eckenfelder form, expresses the ratio as A/S = [1.3 x Sa x (f x P - 1) x R] / (Sa,inf x Q). Here 1.3 is the density of air in mg per mL at about 20 C, converting a volumetric solubility to a mass; Sa is the saturation solubility of air in water at the operating temperature, about 18.7 mL/L at 20 C; f is the fraction of saturation actually achieved in the saturator; P is absolute saturator pressure in atmospheres; R is the pressurised recycle flow; Sa,inf is the influent suspended solids concentration; and Q is the feed flow.

Why the term is (f x P - 1), not P

Henry's law makes the dissolved air load proportional to absolute pressure, so f x P is the air the recycle stream holds under pressure. The minus one subtracts what remains dissolved once the stream is throttled back to atmospheric. Only the difference nucleates as bubbles. That algebraic detail is why pressure delivers diminishing returns: going from three to six atmospheres absolute more than doubles the released air, but each further increment buys proportionally less against the pump energy it costs.

Pressure also sets bubble size. Reported averages are around 82 micron at 200 kPa and around 62 micron at 500 kPa, and DS211 adopts 60 micron in the contact zone and 100 micron in the separation zone for design at roughly 500 kPa. The same guideline warns that raising pressure does not necessarily produce better operation, because high pressure and high temperature both promote coalescence. Bubbles that merge on the way up rise faster, detach from agglomerates and break up flocs that had already formed.

Saturator efficiency is not a rounding factor

The term f carries as much weight in the equation as pressure does, and it is the term most often assumed rather than specified. DS211 puts unpacked saturator efficiency at 50 to 85 per cent; packed saturators are credited with up to 90 per cent, and internal-recycling designs with claims approaching complete saturation. Published values elsewhere put unpacked units at 60 to 70 per cent and packed units near 90. Moving f from 0.6 to 0.9 at the same pressure raises the released air per litre of recycle substantially, which is almost always cheaper than a larger recycle pump. The catch is fouling: DS211 bars packed saturators from waste activated sludge thickening, or any duty where the recycle stream is likely to carry more than 5 mg/L of suspended solids, because the packing blinds.

Why A/S collapses when the feed strengthens

Hold recycle, pressure, saturation fraction and temperature constant and the achieved ratio is inversely proportional to the influent solids concentration. Double the TSS and you halve the ratio. This is the effect operators feel most sharply, because feed strength swings far more, and far faster, than temperature or saturator pressure ever will.

The published worked example is instructive. A unit delivering a fixed air mass and running at A/S = 0.05 on a 600 mg/L feed drops to 0.025 when the feed reaches 1,200 mg/L. Restoring the ratio means doubling the released air, and because released air is linear in recycle flow, that means doubling the recycle, from 30 to 60 per cent in the cited case. The alternative, holding recycle and raising pressure to double the (f x P - 1) term, would demand close to 11 atmospheres absolute, which is rarely practical. Recycle is the usual lever precisely because pressure is not. The design consequence is blunt: size the air system on the peak solids load, not the average.

Temperature: the derate that gets discovered in July

Air solubility falls with temperature and enters the equation directly through Sa. Published values run 28.8 mL/L at 0 C, 22.8 at 10 C, 18.7 at 20 C and 15.7 at 30 C. Between 10 and 30 C the deliverable air per litre of recycle drops by roughly a third at constant pressure, and because A/S is linear in Sa, a saturator sized only for cool operation can undershoot its target by 15 to 20 per cent on a hot day.

This bites hardest exactly where DAF is most needed. Dairy pasteurisation rinse, clean-in-place effluent and rendering streams arrive warm and carry heavy FOG loads, so reduced air availability coincides with peak demand for it. Evaluate the governing equation at the warmest expected operating temperature, and because the solubility constant varies with both temperature and altitude, design for a variable recycle ratio rather than a fixed one.

The target is set by the floc, not the flowsheet

Denser, hydrophobic solids attach air readily and need little of it. Light, hydrophilic, highly hydrated flocs need a great deal of bubble surface per unit mass. Published application bands bracket the design:

• Oil, grease and API-conditioned FOG: roughly 0.005 to 0.02 kg air per kg solids, since hydrophobic low-density material attaches air easily
• Food, beverage and dairy effluent: roughly 0.02 to 0.04, reflecting mixed FOG and protein flocs of moderate demand
• Coagulated and flocculated industrial TSS: roughly 0.03 to 0.05, driven by hydrated metal-hydroxide flocs with high surface demand
• Waste activated sludge thickening: roughly 0.02 to 0.06, with DS211 setting a floor of 0.02 for WAS
• Algae and low-density biomass: roughly 0.04 to 0.06, because the density difference is minimal and abundant microbubbles are required

Use the bands to bracket, then confirm by bench flotation and jar testing on the actual effluent; the optimal ratio cannot be derived from textbook values alone. And none of it works without chemical conditioning. An undersized flocculation chamber produces weak, shear-sensitive flocs that break apart before they reach the contact zone, and no amount of air will rescue them.

Where the air ratio meets the hydraulics

A/S and tank area are not independent variables. DS211 expresses the limiting down-flow velocity in the separation zone as a function of the air-to-solids ratio, vL = K1 x as^K2 - K3, with constants determined empirically per water type: 231, 0.87 and 1.5 m/h for waste activated sludge, for instance. More air raises the hydraulic loading the separation zone can tolerate.

But the recycle stream is itself hydraulic load. The limiting velocity must satisfy vL greater than or equal to Qi x (1 + r) / Asz, so doubling recycle to rescue a collapsing ratio also raises the loading on the separation zone. It is possible to chase your own tail. DS211's guidance is that clarification duty typically runs 6 to 20 per cent recycle at 300 to 750 kPa, while thickening runs 50 to 250 per cent at 400 to 600 kPa, with limiting velocities generally 5 to 15 m/h for clarification and around 6.2 m/h for WAS thickening.

One configuration rule deserves separate emphasis, because it silently destroys air performance the equation says you have already paid for: the recycle must be released into the feed inside the contact zone, never into the feed pipe outside the unit. Coalescence in that pipe produces large, fast-rising bubbles that disrupt floc formation, and DS211 puts the resulting loss of recycle system efficiency as high as 300 per cent.

A sequence that survives commissioning

• Fix the A/S target by bench flotation testing on the real effluent, using published application bands only to bracket the test range
• Compute the solids mass rate at peak influent concentration and peak flow, never at average conditions
• Set the saturator physics: choose the pressure, specify rather than assume the saturation fraction f, and evaluate air solubility at the warmest expected water temperature
• Solve for the recycle ratio, remembering it depends only on the target, the influent concentration and the saturator physics, not on plant size
• Check the separation zone hydraulics at that recycle rate, add margin on both pump and saturator, and verify the achieved ratio on the commissioned unit rather than trusting the datasheet

The judgement to carry away is that a DAF is an air delivery machine with a tank attached. Tank area sets the ceiling on hydraulic loading, but the air-to-solids ratio decides whether there is anything worth separating in the first place. Specify against a tested ratio at peak solids and the warmest expected water, confirm the recycle pump and saturator can still reach it under those conditions, and treat any quoted figure that did not come from your own effluent as a starting hypothesis rather than a design basis. AIRFIN manufactures dissolved air flotation systems in-house, using fine air bubbles to float suspended solids, oils and light contaminants for skimming on food, beverage and distillery effluent, and can advise on A/S targets, recycle sizing and retrofits to existing units.