Why Textile Dyeing Wastewater Breaks the Standard DAF-vs-API Comparison
For high-color, high-TDS reactive dyeing wastewater, a DAF unit is the better first-stage oily-water remover: it delivers 90-98% FOG and 70-90% emulsified oil removal in 10-30 minutes, while an API separator's Stokes-Law efficiency collapses in the 10,000-30,000 mg/L TDS brine typical of a dye bath — its free-oil removal often drops below 50% and emulsified oil stays under 20%. A hybrid of API roughing plus DAF polishing is the most common dyeing-mill configuration in 2026.
Reactive dyeing effluent is not refinery oily water. A dye-bath discharge carries hydrolyzed reactive dyes (the spent Remazol, Procion and Reactive H chromophores that no longer bond to fiber), sizing waxes from desize baths, machine lubricants and antifoaming oils from kier boiler leaks, and surfactant loads from soaping and scouring. Those oils arrive already emulsified at 10-30% of the load, the exact fraction that gravity separators cannot touch. Generic comparison pages that quote 60-75% free-oil removal for API separators are written for refinery desalters, where oil droplets are 100-600 µm and TDS is below 2,000 mg/L. Neither condition holds at a reactive dyeing mill.
The second variable the generic pages skip is TDS. Glauber salt (Na2SO4) and NaCl from reactive baths routinely push dye-house effluent to 10,000-30,000 mg/L TDS. Brine raises water viscosity to roughly 1.05-1.15 cP and narrows the oil-water density difference from about 0.10 g/cm³ in fresh water to 0.05-0.07 g/cm³ in concentrated salt water. Both terms sit in the denominator and numerator of Stokes' Law, so rise velocity — the only mechanism an API separator has — drops by 30-50% before the engineer even runs the hydraulics. The third variable is color. Hydrolyzed reactive dyes are anionic and hydrophilic, so they do not separate on their own, but a cationic coagulant such as polyaluminum chloride (PAC) at 50-200 mg/L plus an anionic polymer can drag 20-40% of that color into the DAF float. The API separator has no chemical stage and removes zero color, so the mill either pays for color removal downstream or accepts a higher MBR loading. None of the top three ranking pages for "DAF vs API separator" addresses any of these three textile-specific realities.
The Physics: How TDS and Color Reshape DAF and API Performance
Stokes' Law sets the rise velocity of an oil droplet as v = (g · Δρ · d²) / (18 · μ), where Δρ is the density difference between oil and the surrounding water, d is droplet diameter, and μ is dynamic viscosity. At TDS 20,000 mg/L and 30 °C, μ rises from 0.80 cP (fresh water) to roughly 0.90 cP — about a 10% increase — and Δρ between a 0.85 g/cm³ machine oil and the brine falls from 0.15 g/cm³ to about 0.10 g/cm³. Putting both into the equation for a 100 µm droplet: fresh-water rise velocity is around 0.10 m/h; in 20,000 mg/L TDS brine it falls to 0.05-0.07 m/h, a 30-50% penalty. An API separator designed for the 60-75% headline removal at 0.5-1.5 m/h surface loading must either be lengthened by 1.5-2× or accept a 100 µm minimum droplet size that the soaping effluent routinely violates. The same penalty does not touch a DAF unit, because DAF uses 10-100 µm bubbles to attach to chemically destabilized oil droplets — the bubble does the lifting, not the oil-brine density difference.
Color behaves differently. Hydrolyzed reactive dyes carry sulfonate and chlorotriazine groups that stay dissolved in the aqueous phase, so they will not rise on their own regardless of the unit. The DAF solves this with chemistry: PAC at 50-200 mg/L neutralizes the anionic charge and forms microflocs; an anionic polyacrylamide at 1-5 mg/L bridges those flocs; the DAF microbubbles then lift the floc to the surface. Field results at Bangladesh and Türkiye reactive mills show 20-40% color removal (measured as Pt-Co absorbance at 436 nm) in the float, before any biological stage. The API separator, by contrast, runs without coagulant and has no float chemistry, so color removal in an API-only train is effectively zero — the mill pays the full color load with MBR or ozone downstream.
Emulsified oil is where the technology gap widens most. Reactive-dyeing soaping effluent typically contains 200-800 mg/L emulsified oil at droplet sizes of 1-20 µm, well below the 100 µm cut-off that Stokes' Law-based separators need. DAF with chemical pretreatment captures 70-90% of this fraction; API captures under 20%. The downstream MBR membrane or SBR biomass then sees an order-of-magnitude lower oil load with a DAF lead unit, which protects the membrane against fouling and keeps the biological stage aerobic.
| Parameter | API separator at TDS 20,000 mg/L | DAF with PAC + polymer |
|---|---|---|
| Oil droplet size handled | ≥100 µm (otherwise rise velocity collapses) | 1-20 µm after chemical destabilization |
| Rise velocity, 100 µm droplet | 0.05-0.07 m/h (vs 0.10 m/h fresh water) | Not applicable — bubble attaches, droplet floats |
| Free oil removal | 40-55% (vs 60-75% headline) | 90-98% |
| Emulsified oil removal | <20% | 70-90% |
| Color co-removal (Pt-Co) | ~0% (no chemical stage) | 20-40% with PAC at 50-200 mg/L |
| TDS sensitivity | Strong — 30-50% loss above 10,000 mg/L | Negligible — chemistry is the driver, not density |
DAF vs API Separator — Textile Dyeing Parameter Comparison
The numbers below are sized for a 2,000 m³/d reactive dyeing mill with TDS 15,000 mg/L, FOG 400 mg/L, emulsified oil 300 mg/L, and Pt-Co color 1,500-3,000. DAF figures draw from the HydropureWater DAF vs API pillar and field installations in 2025-2026; API figures reflect API 421 design rules adjusted for the TDS penalty above.
| Parameter | API separator | DAF (ZSQ series) |
|---|---|---|
| Removal mechanism | Gravity / Stokes' Law | Microbubble flotation (10-100 µm bubbles) + coagulation |
| FOG removal (free + emulsified) | 40-55% (TDS 15,000 mg/L) | 90-98% |
| Emulsified oil removal | <20% | 70-90% |
| TSS removal | 40-60% | 85-95% |
| Effluent O&G | 50-100 mg/L (raw) | 10-20 mg/L |
| Surface loading | 0.5-1.5 m/h | 10-20 m/h |
| Retention time | 1.5-2.5 h | 10-30 min (with chemical aid) |
| Footprint per 100 m³/h | ~250-400 m² (long rectangular tank) | ~30-50 m² (skid-mounted, 30-50% less than API) |
| CAPEX (USD per m³ capacity) | $150-300 (concrete, passive) | $400-800 (steel, compressor, saturator) |
| OPEX (USD per m³ treated) | $0.02-0.05 (minimal) | $0.10-0.20 (PAC + polymer + power) |
| Color co-removal (Pt-Co) | ~0% | 20-40% with PAC 50-200 mg/L |
| Chemical requirement | None (passive) | PAC 50-200 mg/L + anionic polymer 1-5 mg/L |
| TDS tolerance | Degrades 30-50% above 10,000 mg/L | Effectively unchanged (chemistry-driven) |
The two textile-only rows at the bottom are the differentiators that make a generic refinery comparison unusable. A dyeing mill that specifies a standalone API separator at TDS 15,000 mg/L is paying for a 0.5-1.5 m/h surface-loading tank that delivers 40-55% free-oil removal and 0% color knock-down, then handing the residual 50-100 mg/L O&G and full color load to a biological stage that was never sized for it.
Recommended First-Stage Flow Sheet for a 2,000 m³/d Reactive Dyeing Mill
The configuration that 2026 dyeing mills in Bangladesh, India, Türkiye, Vietnam and China are actually running looks like this: rotary bar screen (3-5 mm aperture) to lift lint and fiber → equalization tank sized for 8-12 h HRT to absorb bath-dump peaks → cooling tower or plate heat exchanger to bring the 60-80 °C reactive bath discharge down to ≤40 °C, since DAF biology downstream and most APIs lose efficiency above 45 °C → API roughing separator designed per API 421 to lift bulk free oil and protect the DAF from slugs → ZSQ series industrial DAF system as the polishing unit, fed by a HydropureWater automatic chemical dosing system that meters PAC at 50-200 mg/L and anionic polymer at 1-5 mg/L based on flow-proportional signals → biological stage (MBR or SBR) → sand filter or UF for TSS polishing → optional RO for water reuse back to the dye house.
The API roughing step is the lever most engineers under-use. A small API unit ahead of the DAF catches the 100-600 µm free-oil fraction that would otherwise sit on top of the DAF as a thick rag layer, consuming float capacity and pulling PAC dose up by 20-30%. With the API doing the bulk separation, the DAF is left to do what only DAF can do: capture 1-20 µm emulsified oil and pull 20-40% of the color into the float. The DAF float, by the way, is 3-5% solids — handle it as hazardous sludge and dewater on a small belt press, do not send it back to equalization. The rotary bar screen in front of the equalization tank is the second-most-skipped step on real P&IDs; reactive-dye effluent carries 50-200 mg/L fiber and lint that wraps around DAF skimmer bearings and starves API scrapers, and a rotary mechanical bar screen sized for 3-5 mm aperture is the cheap insurance.
Some 2026 mills skip the API entirely and run screen → equalize → cool → DAF only. That decision is defensible when land is tight (API tanks consume 250-400 m² per 100 m³/h), when influent FOG is below 200 mg/L, or when TDS is above 20,000 mg/L and the API rise-velocity penalty is severe enough that the unit barely earns its footprint. For TDS 10,000-15,000 mg/L and FOG 400-800 mg/L, the hybrid API + DAF is the default and the spec the EPC should carry unless the plant manager overrides it on land or capital grounds.
Cost and Footprint for a Textile Mill ETP in 2026
Working the numbers for a 2,000 m³/d mill (≈ 83 m³/h average, 100 m³/h peak): a passive concrete API roughing unit runs $150-300 per m³ of capacity, or roughly $30,000-60,000 installed for the roughing step. The ZSQ series industrial DAF system in the polishing role sits at $400-800 per m³ capacity, or $80,000-160,000 for the same flow — a 2-3× CAPEX premium that is real but smaller than it looks once footprint enters the equation. A 100 m³/h API needs 250-400 m² of civil space; a 100 m³/h DAF needs 30-50 m². In a Bangladesh or Türkiye urban mill where civil space runs $40-80 per m², the API land cost alone closes $8,800-28,000 of the CAPEX gap, or 20-30% of the differential. Add the avoided emulsion spike risk to a downstream MBR and the DAF premium compresses further.
On the operating side, DAF OPEX is dominated by chemistry: PAC + polymer at $0.05-0.15 per m³ of treated wastewater, plus compressed-air and recirculation-pump power. Total DAF OPEX lands at $0.10-0.20 per m³ including power, maintenance, and skimmer wear. The hybrid configuration lowers DAF chemical dose by 15-25% because the API has already lifted the bulk free oil — for an 83 m³/h plant running 20 hours a day, that is $6,000-12,000 per year in chemistry savings, which over a 10-year DAF life is enough to retire the API CAPEX twice. The real 2026 lever for the engineer defending the spec to a CFO is therefore the chemical line, not the headline CAPEX number.
Selection Checklist for a Textile Mill ETP Engineer
Run this go/no-go before you commit to a P&ID: (1) If TDS is above 10,000 mg/L and the influent carries emulsified oil from soaping or finishing, specify DAF alone, or API + DAF if free oil is also above 200 mg/L. (2) If TDS is below 3,000 mg/L, oil is free, and land is cheap, a standalone API separator per API 421 is acceptable and the cheapest first stage. (3) If the mill is space-constrained or has a strict <20 mg/L O&G discharge or reuse target (e.g., RO feed protection), DAF only. (4) If the mill has 0.5-1 ha of free land and is processing only free oil, API is cheaper — but confirm downstream biology can absorb 50-100 mg/L O&G without losing MLSS. (5) Whatever the unit, size the chemical dosing package to the actual FOG and color load, not a generic 50 mg/L PAC dose; the HydropureWater automatic chemical dosing system is the most common pairing with the ZSQ DAF for reactive dye streams because it tracks flow-proportional setpoints and avoids overdosing during equalization-tank swings. Once those five answers are in writing, the spec is defensible to procurement, to the environmental officer, and to the auditor who will eventually ask why the ETP was built the way it was.
Frequently Asked Questions
Does an API separator work on reactive dyeing wastewater above TDS 10,000 mg/L?
An API separator will still operate, but its free-oil removal falls from the 60-75% headline to roughly 40-55% at TDS 15,000 mg/L, and emulsified oil removal stays under 20%. The drop is set by Stokes' Law: brine viscosity rises about 5-10% and the oil-brine density differential narrows from 0.15 to about 0.10 g/cm³, cutting rise velocity for a 100 µm droplet from 0.10 m/h to 0.05-0.07 m/h. A mill that needs 60-75% free-oil removal at high TDS has to lengthen the API tank by 1.5-2× or add a DAF polishing step downstream.
Can a DAF unit remove color from hydrolyzed reactive dyes?
Yes, but only with chemistry. Hydrolyzed reactive dyes are anionic and hydrophilic, so they will not float on their own. A cationic coagulant such as polyaluminum chloride (PAC) at 50-200 mg/L plus an anionic polyacrylamide at 1-5 mg/L destabilizes the chromophores and drags 20-40% of Pt-Co color into the float. Field data from Bangladesh and Türkiye reactive mills (2025-2026) confirms this range, with the higher end at higher PAC dose and tighter pH control at 7.0-7.5. The remaining 60-80% of color is handled in the MBR or downstream ozone stage.
What is the recommended first-stage flow sheet for a 2,000 m³/d reactive dyeing mill in 2026?
The 2026 default is screen (3-5 mm aperture) → equalization (8-12 h HRT) → cooling (effluent ≤40 °C) → API roughing (per API 421) → DAF polishing with PAC + anionic polymer → MBR or SBR → sand filter or UF → optional RO. Mills with land under 0.2 ha or TDS above 20,000 mg/L may drop the API and run screen → equalize → cool → DAF only, accepting a 15-25% higher DAF chemical dose in exchange for footprint and capital. Both configurations protect the downstream MBR membrane against oil fouling and give the biological stage a 50-100 mg/L O&G influent instead of the 200-400 mg/L it would see with no primary separation.
