Why E-Stage Effluent Is a Different DAF Problem
E-stage wash filtrate — the alkaline extraction-stage filtrate from a chlorine-dioxide bleach plant — is the single most difficult stream a mill will route to a dissolved air flotation unit. It arrives at pH 10–13 and 60–80°C, carrying 200–800 mg/L TSS, 1,500–3,500 mg/L COD, 1,000–3,000 Pt-Co color, and 5–25 mg/L adsorbable organically-bound halogens (AOX). These contaminants exist primarily as sub-100 μm anionic lignin colloids stabilized by the high pH and residual oxidant (ClO₂, HOCl), not as settleable fibers. Direct copy of municipal DAF parameters — hydraulic loading 20–25 m/h, no pH correction, anionic polymer — produces poor TSS removal and excessive AOX carry-over to downstream biology, because the floc density and surface charge of lignin particles differ fundamentally from domestic biosolids or FOG. Treating E-stage as "just another colloidal wastewater" is the most common sizing error a mill engineer makes in 2026, and it is the reason DAF units on bleach-plant service are routinely undersized or mis-chemistried.
| Parameter | Raw E-stage filtrate | DAF feed target (after pre-conditioning) |
|---|---|---|
| pH | 10–13 | 8.5–9.5 |
| Temperature | 60–80°C | < 45°C |
| TSS | 200–800 mg/L | 200–800 mg/L |
| COD | 1,500–3,500 mg/L | 1,500–3,500 mg/L |
| Color | 1,000–3,000 Pt-Co | 1,000–3,000 Pt-Co |
| AOX | 5–25 mg/L | 5–25 mg/L |
| Dominant colloid | Anionic lignin, 0.1–10 μm | Charge-neutralized microfloc, 30–80 μm |
The MDPI review by Bolto & Xie (Processes 7(6):374, 2019) on polymer use in flotation treatment makes the point clearly: the chemistry — charge density, molecular weight, dose — is what determines removal on colloidal streams, not the mechanical size of the vessel. E-stage demands that point be respected at the drawing-board stage.
Step 1 — Characterize the E-Stage Influent
Size a DAF from data, not from a brochure. Pull a representative composite of the E-stage filtrate over at least three shifts and at peak, average, and minimum digester/bleach-line throughput — batch digester blow and continuous bleach-line grade-change events can swing TSS by a factor of three in under an hour (Zhongsheng field data, 2026). The minimum analytical panel is flow (m³/h), pH, temperature, TSS, COD, BOD, color (Pt-Co), AOX, conductivity, and residual oxidant (ClO₂, HOCl, Cl₂). Without residual oxidant, you cannot tell whether the AOX you are measuring is already chlorinated organic matter or active chlorine still free to react downstream — that distinction will set your polymer demand and your vent-gas treatment in Step 4.
Document upstream screening: a 2–5 mm opening rotary bar screen — for example, a GX series rotary bar screen — should already be removing fiber and pulp losses before the DAF. If the screen is missing or oversized, fiber will load the float blanket, swamp the saturator recycle, and look like a "DAF failure" that is actually a pre-treatment failure. Required DAF influent targets after pre-cooling and partial neutralization: pH 8.5–9.5, temperature < 45°C, TSS 200–800 mg/L, COD 1,500–3,500 mg/L. These are not optional — they are the window inside which cationic polyelectrolyte performs on lignin. For comparison, a low-strength stream such as a paper-machine white water is sized quite differently, as covered in DAF sizing for white water discharges.
Step 2 — Set Hydraulic and Solids Loading Rates
Use a hydraulic loading of 10–15 m/h for E-stage. This is lower than the 20–25 m/h used for easily-floated FOG or food-processing streams, and it reflects the higher effective density of lignin flocs (0.95–1.05 g/cm³) versus oil-laden flocs (0.85–0.95 g/cm³). At 15 m/h you still achieve a footprint roughly five times smaller than an equivalent clarifier, but you give the bubble-floc aggregate the residence time it needs to rise. Calculate the required flotation-zone surface area as A = Q / HLR, where Q is the peak hourly flow plus the saturator recycle stream. On a 200 m³/h bleach line at HLR 12 m/h, A is approximately 17 m² of effective flotation surface.
Cap solids loading at 8–12 kg TSS/m²·h. Above 12 kg/m²·h the float blanket loses stability, clarified-turbidity spikes from 5–10 NTU to > 50 NTU, and polymer consumption climbs faster than TSS removed. If peak diurnal flow would force an oversized tank, install a 4–8 h HRT equalization basin upstream — the cost of EQ plus a properly-sized DAF is almost always lower than over-sizing the flotation unit. For comparison, the same HLR logic on a starch-rich food stream produces different saturator settings, as detailed in DAF sizing for potato starch water.
Step 3 — Design the Air-Saturation and Bubble System
Micro-bubble generation is the single most important physical factor in DAF performance on E-stage colloids. Target a bubble diameter of 30–50 μm. Below 20 μm the bubbles lack enough buoyant force to lift a 0.5–1.0 g/cm³ lignin floc, even at full saturation; above 80 μm the bubbles rise too fast and do not attach efficiently. Operate the saturator at 4–6 bar gauge; lower pressure fails to dissolve enough air to reach the design air-to-solids ratio, higher pressure wastes compression energy with diminishing solubility returns on a hot alkaline stream.
Set the air-to-solids ratio (A/S) to 0.02–0.05 g air per g TSS. Use the low end of the range (0.02–0.03) when TSS is above 600 mg/L — there is plenty of particle surface, so excess air only churns the blanket. Use the high end (0.04–0.05) when TSS is below 400 mg/L or when floc density is on the high side of the lignin range. Hold the recycle ratio at 10–30%, with 20% the practical sweet spot for E-stage: high enough to give a clean supernatant (target turbidity < 10 NTU off the flotation zone), low enough to keep the saturator pump and compressor off the mill's largest auxiliary load. These numbers are consistent with the floc-size-and-density framework summarized in the CRC Press chapter on flotation mechanics (DOI 10.1201/9781003073154-6).
Step 4 — Select Coagulant and Polymer Chemistry
Mechanical sizing without chemistry is wasted capital on E-stage. Begin with pre-conditioning: bring pH from 10–13 down to 8.5–9.5 using sulfuric acid (typical dose 1.5–3.0 g/L on E-stage filtrate) or CO₂ if the mill already has a flue-gas CO₂ stream; cool below 45°C with a plate heat exchanger on streams above 60°C — at 70°C polymer hydrolysis is rapid and dose demand roughly doubles. Once conditioned, dose a high-charge-density cationic polyelectrolyte at 5–20 mg/L as the primary flocculant for the anionic lignin colloids. Bolto & Xie (MDPI Processes 7(6):374, 2019) document that cationic polyelectrolytes of medium-to-high molecular weight and high charge density consistently outperform anionic or non-ionic alternatives on colloidal, low-density streams — the same physics that governs E-stage.
For high color or high AOX streams that resist the primary polymer, add a coagulant aid: polyaluminum chloride (PAC) at 20–50 mg/L or ferric chloride at 30–80 mg/L to sweep residual colloidal material and reduce color by an additional 20–30%. Dose the coagulant 30–60 s upstream of the polymer injection point, then the polymer, then provide 3–5 min of slow mixing (G ≈ 50–80 s⁻¹) for floc growth. Jar-test the polymer on actual E-stage filtrate, not a synthetic — bench-scale DAF tests at 1–2 L volume run at the design HLR will predict full-scale TSS and color removal within 5–10%, which is well inside the safety factor for final equipment selection. For automated dose control, specify a PLC-controlled coagulant and polymer dosing skid with flow-paced setpoints and streaming-current feedback.
Step 5 — Size the Vessel, Skimmers, and Sludge Handling
Set the effective hydraulic retention time in the flotation zone to 15–25 minutes — longer (20–25 min) for cooler, denser flocs and shorter (15–18 min) for warm, well-conditioned feeds. Select a surface skimmer with a travel speed of 0.5–1.5 m/min and a beach or scraping ramp rated for the thick, dark, lignin-rich float; E-stage float solids run 3–6% dry solids, which is already dense enough to feed a dewatering device directly without thickening. Pipe the float to a plate and frame filter press for dewatering to 30–35% DS cake — this avoids the odor and re-dissolved-AOX problems of a DAF-thickener-lagoon train, and it sends a handleable solid to boiler fuel or landfill.
Use the surface area from Step 2 to select the closest standard ZSQ series DAF system in the 4–300 m³/h flow range. All ZSQ models accept field-customized saturator sizing, recycle ratios, and inlet/outlet geometry for high-strength streams; for E-stage service, oversize the saturator pump by 10–15% relative to municipal spec to maintain the 4–6 bar pressure drop at the higher recycle temperature.
Step 6 — Controls, Commissioning, and Operating Targets
Continuous online monitoring turns the design into a controllable unit. Install pH, temperature, flow, TSS (solar- or optical-sensor type), and polymer-dose feedback on the PLC — streaming-current or charge-analyzer feedback on the polymer line is the single most effective way to avoid both over- and under-flocculation during diurnal swings. Target DAF effluent: TSS ≤ 50 mg/L, color reduction ≥ 70%, AOX reduction 30–50% (Zhongsheng field data, 2026). These targets protect the downstream biological polishing stage — typically an MBBR or MBR — from chlorinated shock loads, which is the most common cause of biofilm die-off in pulp-mill effluent trains. See the companion guide on MBBR sizing for white water for the biological-side numbers.
Common 2026 commissioning faults: bubbles too large (check saturator pressure and look for air leaks on the recycle line), floc breaking up in the flotation zone (recheck polymer dose and mixing G), and sludge carry-over into the launder (lower HLR by 1–2 m/h or check scraper speed). For mills facing EU Best Available Techniques (BAT-AEL) discharge limits for AOX and color, DAF alone is insufficient — pair with MBR or, for zero-liquid-discharge sites, with evaporation/crystallization polishing.
DAF Equipment Selection Matrix for E-Stage Effluent
The table below maps typical E-stage flow rates to a ZSQ series model, footprint, saturator pump power, and the polymer dose range that bench testing on bleach E-stage has validated (Zhongsheng field data, 2026). All sizes are field-customizable on hydraulic loading and saturator capacity for high-strength E-stage streams; when peak flow sits between two standard models, size to the next-larger unit rather than over-stretching the smaller one.
| ZSQ model | Design flow (m³/h) | Footprint (m²) | Saturator pump (kW) | Cationic polymer dose (mg/L) | Expected TSS / color / AOX removal |
|---|---|---|---|---|---|
| ZSQ-5 | 4–8 | 2–3 | 1.5–2.2 | 10–20 | 80–90% / 70–80% / 30–40% |
| ZSQ-20 | 15–25 | 4–6 | 3.0–4.0 | 8–18 | 80–90% / 70–80% / 30–40% |
| ZSQ-50 | 40–60 | 8–12 | 5.5–7.5 | 6–15 | 80–90% / 70–85% / 35–45% |
| ZSQ-100 | 80–120 | 14–20 | 11–15 | 5–12 | 85–92% / 75–85% / 35–50% |
| ZSQ-200 | 160–240 | 25–35 | 18–22 | 5–10 | 85–92% / 75–85% / 40–50% |
| ZSQ-300 | 260–340 | 38–50 | 30–37 | 5–10 | 85–92% / 75–85% / 40–50% |
Confirm final selection with a bench-scale DAF test on the mill's actual E-stage filtrate — the dose numbers in the table are typical, but local lignin origin (hardwood vs. softwood), kappa number, and bleach sequence can shift the optimum by 30–50%.
Frequently Asked Questions
What removal efficiencies can a DAF achieve on bleach E-stage effluent?
On a properly conditioned feed (pH 8.5–9.5, < 45°C, with cationic polymer at 5–20 mg/L), a DAF on E-stage filtrate will deliver 80–92% TSS removal, 70–85% color reduction, and 30–50% AOX reduction (Zhongsheng field data, 2026). These numbers assume the lignin is the dominant colloidal load; residual AOX in true solution passes through the DAF and must be handled in the downstream biological stage.
What type of polymer works best for E-stage?
A high-charge-density cationic polyelectrolyte at 5–20 mg/L is the primary flocculant for the anionic lignin colloids characteristic of E-stage. For high color or AOX streams, a coagulant aid such as polyaluminum chloride (PAC, 20–50 mg/L) or ferric chloride (30–80 mg/L) improves residual colloidal sweep and adds 20–30% to color removal. Bench-test on actual E-stage filtrate before committing to a dose — see the polymer discussion in Bolto & Xie, MDPI Processes 7(6):374, 2019.
Why is pH adjustment necessary before DAF on E-stage?
Raw E-stage runs at pH 10–13, which keeps lignin as stable, highly charged anionic colloids that resist flocculation. Bringing pH to 8.5–9.5 with sulfuric acid or CO₂ reduces the surface charge and lets the cationic polymer bridge and sweep the colloids into a floatable floc. Without this correction, polymer demand roughly doubles and clarified-turbidity rises by a factor of three to five.
Why route E-stage to a DAF at all instead of sedimentation?
Sedimentation on hot, low-density lignin colloids at pH 10–13 produces overflow turbidity of 100–200 NTU and removes less than 30% of TSS, because the particles are near or below water density. DAF on the same stream, after pH correction to 8.5–9.5 and cationic flocculation, removes 80–92% of TSS at a footprint roughly five times smaller than an equivalent clarifier (Zhongsheng field data, 2026). The flotation step also protects the downstream MBBR or MBR from chlorinated shock loads.
Does a DAF replace biological treatment for E-stage?
No. A DAF removes 30–50% of AOX and 70–85% of color, but mill-scale compliance with EU BAT-AEL or equivalent limits usually requires a biological polishing stage — typically an MBBR for BOD/COD reduction, or an MBR where effluent reuse is the target. The DAF's role is to protect that biology by removing colloidal load, suspended solids, and a fraction of the AOX upfront.
