Coagulation/flocculation, activated carbon, and membranes can each remove roughly 90%+ of color from pulp and paper effluent — but they target different chromophores and carry very different operating costs and risks.
Pulp mill effluent is famously brown — a color driven by lignin and related organics such as chlorolignins and tannins — and mills face growing pressure to decolorize for reuse or discharge. Color is typically measured in Pt‑Co (platinum‑cobalt) or ADMI units (both are standard color scales used in water testing). Three physico‑chemical strategies dominate: coagulation/flocculation (C/F), activated carbon adsorption, and membrane filtration. Quantitatively, coagulation can remove roughly 90–96% of color, activated carbon about 95–97%, and membranes (UF/NF/RO) on the order of 90+% — with cost estimates spanning about $0.2–2 USD per m³ depending on process and region (www.mdpi.com) (www.mdpi.com).
Colored organics and measurement
The bulk of color arises from high‑molecular‑weight lignin fragments and related chromophores; large lignins (about 1,000–8,000 molecular weight) are comparatively easy to remove via coagulation or size‑exclusion steps (nepis.epa.gov). Mills targeting reuse quality routinely pair primary solids removal with targeted color treatment; primary operations can include screens and sedimentation that sit upstream of chemical or membrane units. For solids control before chemistry, facilities often deploy compact clarification steps; a lamella approach is one option when footprint is constrained, in the same treatment train as a conventional clarifier.
Chemical coagulation/flocculation mechanisms
C/F destabilizes colloids and precipitates high‑molecular‑weight organics so they can settle; flocculation then aggregates the destabilized particles. In practice this step excels against colloidal lignin polymers and tannins.
Effectiveness is high in controlled trials. An alum dose of 1.2 g/L at pH≈6 achieved 96% color reduction in primary‑treated pulp mill effluent (www.mdpi.com). Lime (Ca(OH)₂) at ~1.4 g/L, pH≈11, removed ~91.6% (www.mdpi.com), while MgSO₄ (3.4 g/L) removed ~89% (www.mdpi.com). Turbidity and TSS were also reduced (>90% in these trials; same sources).
Optimum pH matters. Alum performed best near pH≈6; lime and MgSO₄ required alkaline conditions (roughly pH 9–11) (www.mdpi.com) (www.mdpi.com). Doses commonly fall in the 1–3 g/L range, with primary sedimentation ahead of C/F to remove large suspended solids.
Chemical metering is standardized; mills often pair coagulant addition with a dedicated dosing pump to hold pH and doses at their setpoints without operator drift.
Costs and residuals are material. A full‑scale physicochemical scheme (settling + coagulation + adsorption) has been estimated at about $2.09/m³ of wastewater (www.mdpi.com). Coagulant addition generates sludge — up to ~0.16 L per m³ at 1.2 g/L alum in one test (www.mdpi.com) — which requires downstream handling. Coagulants are relatively low‑cost on a per‑kg basis ($0.2–0.5 USD/kg for alum), but required dosages make OPEX nontrivial. In one hybrid study, an integrated coagulation–adsorption process cost ~6.78 Baht (~$0.20)/m³ (ph02.tci-thaijo.org). For color programs that add floc aids, mills incorporate flocculants after coagulants to improve settling.
Bottom line: C/F is well‑proven and scalable. It removes bulk color from colloidal lignins and suspended organics, but residual color often stems from smaller dissolved chromophores that pass through. It is frequently paired with a polishing step. Where footprint is tight, compact settlers can support C/F; lamella plates are one option within conventional primary trains alongside wastewater physical separation hardware.
Activated carbon adsorption performance
Activated carbon (AC) — granular or powdered — adsorbs dissolved organics, including hydrophobic aromatics, dyes, and small lignin fragments. In bench‑scale tests, color and lignin removal reach 94–97% at moderate AC doses. One study using powdered AC made from plastic waste reported 96.5% color and 94.3% lignin removal at 7 h contact, 1.25 g/100 mL, pH 8 (link.springer.com). Another achieved 97% color removal with 2 g/L AC in 1 h (www.icontrolpollution.com). Typical isotherms follow Langmuir–Freundlich behavior, with equilibrium in ~1–7 hours depending on agitation and dose (same sources).
Operating windows are wide: AC doses are often 1–5 g/L (1000–5000 mg/L) with contact times of 1–4 h, and much of the removal can occur in the first hour (www.icontrolpollution.com). AC is typically used after C/F to target residual, dissolved color, or as a standalone step for concentrated streams. For polishing duty, mills source media such as activated carbon designed for organics and color removal.
Cost and handling are consequential. AC is expensive (on the order of $0.5–2.0 per kg) and must be regenerated or replaced; a large mill might use kilograms per m³ of effluent — e.g., 5 kg/m³ at optimum conditions in one estimate (www.mdpi.com). Spent carbon disposal or thermal reactivation adds cost. In one analysis, coagulation plus activated carbon yielded OPEX of about $2.31/m³ (www.mdpi.com). Some tropical studies report lower costs (≈$0.2/m³), likely reflecting lower labor/chemical prices and simpler designs (ph02.tci-thaijo.org).
Scope and limits: AC is versatile and effective, but it does not reduce TSS or inorganic load, and performance declines as carbon saturates. On extremely high‑color streams (e.g., ~20,000 Pt‑Co from bleach filtrate), AC alone can require large volumes (nepis.epa.gov). AC pairs well with upstream C/F, which removes particulates and large lignins so AC can focus on smaller chromophores.
Membrane filtration: UF, NF, and RO
Membranes separate by size and charge. Ultrafiltration (UF, nominal molecular‑weight cutoff ~1–100 kDa), nanofiltration (NF, ~200–1000 Da), and reverse osmosis (RO, which retains virtually all dissolved organics) all target color bodies physically. A 1980 EPA pilot on kraft bleach effluent reported UF would reduce overall color by ~91%, and that UF retained 97–99% of color chromophores on a concentrate basis (nepis.epa.gov). Modern NF/RO systems are tighter, often achieving >95% removal of COD and color in bench tests (>90% COD/color with NF on papermill effluent; same source). Permeate is typically very clear and suitable for reuse.
Operations require discipline. Pulp waters foul; high suspended solids or colloids can quickly blind membranes, so pretreatment (screening, settling, C/F) is standard ahead of membranes. Typical operating pressures: UF ~2–5 bar, NF ~10–30 bar, RO 25–60 bar; energy can run ~0.5–2 kWh per m³. UF primarily removes high‑MW lignin fragments (about 4,000–8,000 MW) (nepis.epa.gov).
Membrane costs are front‑loaded: plants can cost millions of USD, membrane replacements every ~3–5 years add to OPEX, and energy is significant; membranes are often reserved for polishing or reuse loops. Comparative studies suggest membranes become attractive versus advanced oxidation when very high quality reuse or zero‑liquid‑discharge is needed. There is no universal cost‑per‑m³; operational data indicate several $ per m³ including amortized capital (same source). By contrast, single‑stage UF used decades ago was judged “attractive vis‑à‑vis alternative technologies” (nepis.epa.gov).
Hardware choices mirror that spectrum. As pretreatment, mills deploy screening before membranes — for continuous debris control, an automatic screen reduces carryover to downstream units. For pressure‑driven separation, engineers specify ultrafiltration to knock out colloids, step up to nanofiltration for small chromophores, and use brackish water RO when near‑zero color is required. Integrated packages such as membrane systems align with polishing or reuse ambitions. Fouling from oils, pitch, and biocides is a real constraint; continuous cleaning or CIP is common, membranes do not remove free oil or inert solids, and concentrate management is required.
Comparative performance and economics
Removal efficacy: C/F can remove the bulk of colloidal color (often >90%; www.mdpi.com) (www.mdpi.com) but typically leaves some dissolved color. AC and membranes address residual chromophores. Combined C/F + AC can yield essentially colorless water — e.g., <37 Pt‑Co units, “no visible color” (www.mdpi.com). In multiple trials, a sequence of settling + alum coagulation (1.2 g/L) + granular AC delivered final color below 50 Pt‑Co (same source). UF/NF can single‑handedly remove >90% of color but require meticulous pretreatment to avoid fouling.
Costs and economics: Coag/AC processes concentrate cost in chemicals and sludge handling. Reported chemical costs cluster around ~$2–2.3/m³ for coagulation plus adsorption (www.mdpi.com). In a Thailand pilot, a downstream coag+AC step was estimated at ~$0.20/m³ (ph02.tci-thaijo.org). Membranes carry high fixed costs; typical municipal RO plants cost several $ per m³ in developing‑world settings. For a pulp mill, one analysis suggests tertiary AC polishing alone cost ~$1–1.5/m³ (www.mdpi.com).
Suitability by compound type:
- High‑MW lignin/tannin: coagulation is effective (sediments them), AC also adsorbs them, and UF removes them (nepis.epa.gov).
- Low‑MW chromophores/dyes: coagulation is less effective; AC is good; NF/RO are best.
- Inorganic colorants/iron: not typical in pulp effluent; if present, coagulation (e.g., with PACl) can precipitate metals; AC has little effect on metals.
Regulatory context: Most standards emphasize COD, BOD, and TSS rather than explicit color limits; for example, Indonesia’s discharge limits specify COD/TSS but not color (www.mdpi.com). Nonetheless, mills aiming for reuse or aesthetic discharge often target <100 Pt‑Co, and achieving low color usually tracks with lower organic load.
Data‑backed outcomes and design math
In one pilot, alum coagulation (1.2 g/L) cut effluent color 96% (www.mdpi.com). Another bench test showed 97% color removal with activated carbon (2 g/L) in 1 h (www.icontrolpollution.com). An EPA report found a UF system negated ~91% of bleach‑plant color, with 97–99% retention of chromophores on a concentrate basis (nepis.epa.gov).
Chemical costs for these processes often range $0.2–2/m³ (www.mdpi.com) (ph02.tci-thaijo.org). Consider a large mill discharging 10,000 m³/d with initial color 5,000 Pt‑Co: achieving 95% removal (to ~250 Pt‑Co) via coagulation alone may be feasible at roughly ($2/m³)×10,000 m³/d = $20,000/day in chemicals (www.mdpi.com). Adding an AC polishing step (another ~$1–2/m³) or NF would increase costs. Conversely, a target of 80% color reduction could use lower coagulant dosage at a fraction of that cost.
Design implications and product fit
All three options significantly reduce color, but each fits a different place in the train. C/F is the lowest‑complexity, lowest‑barrier step (removal ~90–95% under optimal conditions; www.mdpi.com) (www.mdpi.com) and is often followed by AC for residual color. Membranes can deliver virtually colorless permeate for reuse but with capex, energy, fouling, and concentrate management to plan for. Many mills combine methods — e.g., settling + C/F + AC (and sometimes biological polishing) — to balance cost and effluent targets (www.mdpi.com) (nepis.epa.gov).
The choice is ultimately about targets. If near‑complete decolorization is needed — for example, reuse as boiler feed or meeting strict aesthetic limits — the extra OPEX for AC or NF/RO can be justified; if only regulatory compliance on COD is mandatory, simpler coagulation (or biological treatment) may suffice. In coag trains, linking chemistry to a robust coagulants supply program and ensuring stable pretreatment hydraulics can stabilize both color removal and costs.
