Four complete calculations, from raw production figures to expected effluent quality. These are anonymised composites of real projects, not client case studies — the numbers are representative and the method is exactly the one we use.
The method, in five steps
1. Establish balanced design flow (m³/h) and peak. 2. Calculate solids load = flow × TSS, plus chemical floc from the coagulant dose. 3. Flotation area = design flow ÷ surface loading rate. 4. Air demand = solids load × A/S ratio. 5. Recycle flow = air demand ÷ available dissolved air at working pressure and actual temperature. Step five is where most designs go wrong, because the temperature correction is skipped.
The decisive figure is the 75 % soluble COD. Flotation takes total COD from 2,800 to about 2,100 mg/L — a 25 % reduction. If the consent is 1,000 mg/L COD, the DAF cannot deliver it and anaerobic or aerobic treatment is mandatory. The DAF still earns its place by protecting that biology and cutting its solids load by 90 %.
Example 2: abattoir, 150 cattle/day
Step
Calculation
Result
Throughput
150 cattle/day, 1,200 L effluent per head
180 m³/day
Design flow
Slaughter 8 h, balanced over 16 h
11.3 m³/h, peak 24 m³/h
Influent (blood segregated)
TSS 1,800 mg/L, FOG 900 mg/L, COD 4,500 mg/L
—
Solids load
11.3 × (1.800 + 0.900)
30.5 kg/h
Chemical floc
FeCl₃ 300 mg/L
+2.4 kg/h
Flotation area
11.3 ÷ 6 m/h
1.9 m² → select 2.2 m²
Air demand
32.9 × A/S 0.04
1.32 kg/h
Available air
5.5 bar, 30 °C, 80 % efficient
0.097 kg/m³
Recycle flow
1.32 ÷ 0.097
13.6 m³/h (120 %)
Float sludge
~33 kg/h dry at 5 %
0.66 m³/h, 10.6 m³/day
Expected effluent
—
TSS 60–180 mg/L, FOG 10–30 mg/L, COD 1,600–2,700 mg/L
The 120 % recycle ratio is not an error. High FOG demands a high air-to-solids ratio, and air demand drives recycle flow directly. A supplier quoting 20 % recycle on this duty has sized the unit on a municipal assumption and it will produce a thin float from day one. Note also that blood must be segregated and rendered separately: unsegregated blood roughly triples the COD load and no DAF will compensate.
Example 3: edible oil refinery, 40 t/day
Step
Calculation
Result
Throughput
40 t/day crude, 1.8 m³ effluent per tonne
72 m³/day
Design flow
Balanced over 20 h
3.6 m³/h, peak 6 m³/h
Influent after interception
FOG 1,100 mg/L, TSS 700 mg/L, COD 6,000 mg/L
—
Solids load
3.6 × (1.100 + 0.700)
6.5 kg/h
Chemical floc
PAC 350 mg/L
+0.44 kg/h
Flotation area
3.6 ÷ 5.5 m/h
0.65 m² → select 0.8 m²
Air demand
6.9 × A/S 0.055
0.38 kg/h
Available air
5.5 bar, 35 °C, 80 % efficient
0.090 kg/m³
Recycle flow
0.38 ÷ 0.090
4.2 m³/h (117 %)
Oil recovered
~6.5 kg/h float, 25 % oil, 300 days
~11.7 t/year acid oil
Expected effluent
—
FOG 20–60 mg/L, TSS 50–100 mg/L, COD 2,000–3,500 mg/L
Small flow, high concentration, high recycle. The recovered oil is worth having: at 11.7 t/year and typical biodiesel feedstock pricing, it substantially offsets the chemical and power bill for the whole effluent plant.
Example 4: municipal works, 12,000 m³/day
Step
Calculation
Result
Average flow
12,000 m³/day
500 m³/h
Design flow
Peak factor 1.8
900 m³/h
Influent
TSS 320 mg/L, BOD 280 mg/L, TP 8 mg/L
—
Solids load at peak
900 × 0.320
288 kg/h
Chemical floc
FeCl₃ 70 mg/L
+30 kg/h
Flotation area
900 ÷ 22 m/h high-rate
41 m² → 2 cells × 21 m²
Air demand
318 × A/S 0.012
3.8 kg/h
Available air
5 bar, 25 °C, 80 % efficient
0.103 kg/m³
Recycle flow
3.8 ÷ 0.103
37 m³/h (4 %)
Sludge
~250 kg/h dry at 4 %
6.3 m³/h
Expected effluent
—
TSS 40–65 mg/L, BOD 120–150 mg/L, TP <2 mg/L
The 4 % recycle ratio at the municipal end of the range is the mirror image of the abattoir’s 120 %. Both are correct for their stream, which is precisely why recycle ratio must be calculated rather than assumed.
What these four show
Brewery
Abattoir
Edible oil
Municipal
Design flow
31 m³/h
11.3 m³/h
3.6 m³/h
900 m³/h
Surface loading
8 m/h
6 m/h
5.5 m/h
22 m/h
A/S ratio
0.020
0.040
0.055
0.012
Recycle ratio
13 %
120 %
117 %
4 %
COD removal
25 %
50 %
55 %
50 %
DAF alone meets consent?
No
Usually not
No
As primary stage, yes
Recycle ratio spans thirty-fold across these four cases. Any rule of thumb that quotes a single figure is wrong for at least three of them.
