DAF Energy Consumption and Optimisation: Where the Power Goes

Where does the energy go in a DAF?

A dissolved air flotation unit has a small number of energy consumers, and their ranking is remarkably consistent across installations. In descending order of typical draw:

• Recycle (pressurisation) pump — pumps the recycle stream from near-atmospheric to saturator pressure (4–6 bar). This is almost always the dominant load, often 60–80% of DAF electrical demand.
• Air supply — a small compressor or an eductor/venturi that introduces air into the saturator. Modest, but non-trivial if a dedicated compressor is used.
• Sludge / float scraper drive — a low-power gearmotor, typically well under 1 kW.
• Chemical dosing pumps and mixers — coagulant/polymer metering, usually a few hundred watts each.

Because the recycle pump dwarfs everything else, energy optimisation of a DAF is largely the optimisation of the pressurised recycle loop. Get the recycle ratio and saturator pressure right and you have addressed most of the running cost.

How do you calculate the recycle pump power?

The hydraulic power delivered to a fluid is set by the flow and the pressure rise; the electrical power drawn is that divided by the pump and motor efficiencies. This is the governing relationship for the dominant DAF load.

It is often cleaner to work directly in pressure terms. Since ρ g Q_r H = Q_r × ΔP, the electrical power is simply P = Q_r ΔP / η with Q_r in m³/s and ΔP in Pa (1 bar = 10⁵ Pa). Dividing by the treated flow gives the specific energy in kWh/m³, the metric that matters for benchmarking. See our DAF recycle system design guide for how the recycle ratio itself is set.

Worked example 1: recycle-pump specific energy

Take a base-case industrial DAF treating a feed of Q = 100 m³/h, with a recycle ratio R = 50% (so Q_r = 50 m³/h), a saturator gauge pressure of 6 bar, and a combined pump×motor efficiency η = 0.65.

• Recycle flow: Q_r = 50 m³/h = 0.01389 m³/s.
• Pressure rise: ΔP = 6 bar = 6 × 10⁵ Pa (we neglect the small atmospheric suction offset).
• Hydraulic power: Q_r × ΔP = 0.01389 × 600000 = 8333 W = 8.33 kW.
• Electrical power: P = 8.33 / 0.65 = 12.8 kW.
• Specific energy: divide by the treated flow, not the recycle: 12.8 kW / 100 m³/h = 0.128 kWh/m³.

That single figure — 0.13 kWh/m³ from the recycle pump alone — sits at the upper end of the typical DAF band and immediately flags the 6 bar saturator as the thing worth challenging. Adding, say, 1.5 kW for the air compressor and scrapers lifts the total to roughly 0.14–0.16 kWh/m³.

The air compressor and eductor load

Air must be delivered into the saturator at (or above) saturator pressure. Two arrangements dominate:

• Dedicated air compressor — supplies a small, controlled air flow at 6–7 bar. The isothermal compression power for a gas flow Q_a from p₁ to p₂ is P_comp = p₁ Q_a ln(p₂/p₁) / η_c. Because the air mass is tiny (kg/h, not tonnes/h), this is typically only 0.3–1.5 kW.
• Eductor / venturi injection — draws air by suction on the pressurised recycle. It uses no separate motor, but it imposes a head loss on the recycle pump, so its "cost" is folded back into the pump duty rather than appearing as a separate load.

The practical point: the compressor is a rounding error next to the recycle pump. Optimisation effort belongs on the pump and the pressure it works against, not on shaving compressor kilowatts.

How do pressure and recycle ratio drive energy?

Both levers act linearly on pump power — P ∝ Q_r × ΔP — so cutting either looks attractive. The catch is that both also set the available dissolved air, which must still satisfy the air-to-solids (A/S) requirement. The mass of air released per unit treated flow follows Henry's law:

Herein lies the trade-off. Energy is roughly proportional to R × P, but released air is roughly proportional to R × (P − 1). At high pressure the two nearly track each other, so trimming pressure barely dents the air-per-joule efficiency. At low pressure the −1 term bites: released air falls faster than energy, so going too low is inefficient. There is a genuine optimum, and it usually sits a little below the pressure most units are set to. The saturator and nozzle configuration fixes where that optimum lands.

Worked example 2: cutting saturator pressure 6→5 bar

Return to the base case (Q = 100 m³/h, R = 50%, η = 0.65). Suppose float testing has fixed the required air mass; we drop saturator pressure from 6 to 5 bar and must hold the same delivered air by increasing recycle. Use absolute pressures (P_abs = gauge + 1) and f = 0.9 in the air term.

• Air per unit recycle at 6 bar: (f·P_abs − 1) = (0.9 × 7 − 1) = 5.30 (bar-equivalent).
• Air per unit recycle at 5 bar: (0.9 × 6 − 1) = 4.40.
• Recycle needed to hold air constant: R must rise by 5.30/4.40 = 1.205×, so Q_r goes from 50 to 60.3 m³/h (R = 60%).
• Original pump power (6 bar): from Example 1, 12.8 kW.
• New pump power (5 bar, higher recycle): P = Q_r ΔP / η = (60.3/3600) × 5×10⁵ / 0.65 = 0.01675 × 500000 / 0.65 = 12.9 kW.

The net effect is essentially zero — the pressure saving is almost entirely clawed back by the extra recycle needed to keep the air on target. This is the crucial, counter-intuitive result: when the A/S requirement is binding, cutting saturator pressure alone does not save energy. Real savings come only when the incumbent design carries genuine surplus air (over-recycling or over-pressuring beyond the tested A/S), which is why an energy audit must start from the actual float-test air demand, not the nameplate.

VSD control, turndown and diurnal load

Most industrial effluent flows swing over the day and week. A DAF sized for peak flow but run at a fixed recycle pump speed wastes energy whenever the plant is below peak. A variable-speed drive (VSD) on the recycle pump lets you track the air demand down with flow.

Two cautions temper the affinity-law intuition. First, the recycle pump works against a largely fixed saturator pressure, so head does not fall as the square of speed — the cubic "P ∝ N³" rule that applies to a friction-dominated system is optimistic here. Savings track roughly linearly with recycle turndown at constant pressure. Second, dropping recycle reduces available air, so turndown is bounded by the A/S floor. In practice a VSD still yields meaningful savings by matching recycle to a varying solids load, and by enabling a soft ramp instead of throttling a valve. See how the saturator design sets the pressure the drive must overcome.

Worked example 3: annual cost and VSD payback

Take the base-case recycle pump at 12.8 kW, running 8,760 h/yr at an industrial tariff of £0.20/kWh.

• Annual energy: 12.8 kW × 8760 h = 112,100 kWh/yr.
• Annual cost: 112,100 × £0.20 = £22,400/yr.
• VSD saving: assume the plant averages 70% of peak flow, and a VSD lets recycle track load, giving a conservative 20% average reduction in recycle-pump energy. Saving = 0.20 × £22,400 = £4,480/yr.
• VSD capital + install: for a motor of this size, roughly £6,000–9,000. Take £7,500.
• Simple payback: £7,500 / £4,480 = 1.7 years.

A payback under two years is typical for a continuously running recycle pump with real flow variability, and the drive also cuts mechanical stress from valve throttling and start/stop cycling. On a flat, always-at-peak duty the case weakens — the VSD saves little if there is nothing to turn down.

How does DAF energy compare with alternatives?

DAF is not the cheapest clarifier to run, but it buys a small footprint and high-rate solids/FOG removal. Indicative specific-energy bands for comparable clarification duties:

The comparison also explains why recycle (part-flow) pressurisation won out over full-flow: pressurising only the 30–60% recycle, rather than the entire feed, is the single biggest energy design decision in a DAF. Everything in this article optimises within that already-efficient choice.