Pulp and paper mills that can dial in coagulation–flocculation on the bench routinely drive raw river water to sub‑NTU clarity at scale. The playbook is jar tests, smart chemistry, and patient clarification.
Pulp and paper manufacturing gulps fresh water — on the order of ~60 m³ per tonne of paper (researchgate.net). Raw surface sources swing with the seasons, bringing high turbidity and organic colloids that change without notice (nyjxxb.net).
The default response is coagulation–flocculation (C/F, particle destabilization followed by aggregation) using hydrolyzing metal salts such as alum or ferric chloride/sulfate, or pre‑polymerized options like PAC (polyaluminum chloride), often aided by a high‑molecular‑weight polymer “flocculant” for bridging. Under tuned conditions, C/F alone strips out on the order of 90–99% of turbidity in high‑turbidity surface waters (researchgate.net; link.springer.com), with studies reporting ~95% turbidity removal using ferric chloride and ~~99% with alum under jar‑test conditions (researchgate.net).
The catch: every source behaves differently, so jar testing decides the winner. On Indonesian reservoir raw water at ≈320 mg/L suspended solids, ferric sulfate optimized at ~100 mg/L and pH≈9 (jurnal.ugm.ac.id). In muddy water around ≈250 NTU (NTU, a standard measure of cloudiness), alum dialed in at ~250 mg/L and pH~6 (link.springer.com). Recent head‑to‑heads found all eight PAC formulations tested beat alum on residual turbidity (researchgate.net).
Once floc forms, a clarifier hands off to rapid filtration, pushing clarified water to <0.2–1 NTU — and often to <0.2–0.5 NTU, i.e., drinking‑water clarity — when it reaches the filters (researchgate.net; researchgate.net).
Jar testing: dose, pH, and shear
Jar testing — bench‑scale beaker trials that simulate plant mixing and settling — is the standard way to select coagulant chemistry and polymer aid. It starts with representative raw‑water sampling and baseline measurements (turbidity, pH, alkalinity), then parallel beakers are run through a matrix of dose and pH conditions.
Setups commonly load beakers with different alum, ferric chloride/sulfate, or PAC doses (e.g., 0–300 mg/L). A rapid mix at about 250 rpm for 2 min disperses the coagulant and triggers hydrolysis, followed by flocculation at 40–60 rpm for 10–15 min to grow flocs (studylib.net). After mixing, jars sit undisturbed for 30–60 min to let flocs settle; the supernatant turbidity (or suspended solids) from each jar is then measured.
Polymers are tested in a second pass. Typical cationic polyacrylamide flocculant doses run ~1–5 mg/L, added after 1–2 min of rapid mix; many protocols apply a brief ~10 s pulse at 250 rpm on polymer addition, then revert to slow mixing (studylib.net). The goal is visibly strong, fast‑settling floc. Operators typically record all jar conditions and plot dose–response curves to pinpoint optima.
In practice, dose curves often drop sharply to a broad minimum before rebounding with overdosing (charge reversal). For example, 50–100 mg/L alum may remove ~80–95% turbidity, while 200+ mg/L can reach ~99% (researchgate.net; link.springer.com). PAC frequently shifts the optimum lower — in many cases 5–20 mg/L Al from PAC equals or beats 50–100 mg/L alum — and one case showed ~30% better settled turbidity than alum at the same dose (researchgate.net).
Some plants codify the outcomes into full‑scale setpoints and inventory plans — a move made easier with accurate chemical feed gear such as a dosing pump and standardized reagent stocks. Many mills source base chemistry via coagulants, flocculants, and PAC options like polyaluminum chloride.
Clarification and dual‑media filtration
After dosing, flocculated water enters a clarifier (sedimentation basin) with typical retention time of 1–2 h. Well‑formed floc allows 70–90% removal of remaining suspended solids within the clarifier, and by the outlet, most particles >10–20 µm have settled out. Plants commonly use a clarifier as the settling workhorse before filtration.
The polish comes from rapid multimedia filtration — sand/anthracite beds that screen down to about 5–10 µm equivalents in practice — pushing turbidity to <0.5–1 NTU for process water and often to <0.2–0.5 NTU in drinking‑water applications (researchgate.net; researchgate.net). Dual‑media beds are often built with sand media and a top layer of anthracite; periodic backwash is required to remove accumulated sludge.
Some facilities swap sedimentation for dissolved‑air flotation (DAF) or adopt membrane filtration for clarification; in those cases, pretreatment membranes such as ultrafiltration can play the clarifying role. Conventional sedimentation plus media filtration remains the typical configuration. Overall, a well‑optimized C/F + clarifier + filter train can strip >5 log units of particulate matter and deliver >99% combined turbidity removal with effluent turbidity ≪1 NTU (researchgate.net; researchgate.net).
Choice of coagulant affects sludge volumes and handling: Al‑salts generally produce more voluminous Al(OH)₃ sludge than Fe‑salts or PAC at equivalent metal dose. Higher polymer coagulant‑aid dosing tends to form denser floc, reducing sludge volume. Routine monitoring of turbidity, pH, alkalinity, and residual Al/Fe in the clarified stream anchors process control.
Bench data and source‑specific optima
Jar‑test data points illustrate how conditions shift by source. On “brown water” near 250 NTU, alum at about 0.25 g/L and pH 6 removed most turbidity within 30 min (link.springer.com). On a reservoir feed around 320 mg/L suspended solids, ferric sulfate at ~100 mg/L optimized at pH≈9 (jurnal.ugm.ac.id).
PAC tends to meet targets at lower metal doses: numerous cases show 5–20 mg/L Al from PAC matching or surpassing 50–100 mg/L alum, with one study delivering ~30% better settled turbidity than alum at the same dose (researchgate.net). In a PAC optimization example, the best coagulant (Hyperlon 4064) drove final filtered turbidity to ≈0.15 NTU at ~2 mg/L Al dose (researchgate.net).
Polymer flocculants have a narrow sweet spot: underdosing yields weak, pin‑sized floc while overdosing can restabilize particles via charge effects (studylib.net). Plants typically record jar‑test conditions in spreadsheets, converting results to kg/m³ dose rates that inform dosing setpoints and chemical inventories.
Troubleshooting the clarification line
Raw water variability is the recurring disruptor. When turbidity or organics spike, jar‑test on the day’s water and adjust dose or pH as indicated; manuals recommend increasing coagulant and, if the raw pH falls, adding alkalinity (e.g., lime or caustic) to restore performance (pdfcoffee.com).
If floc will not form, likely causes include too little coagulant, pH/alkalinity mismatch, inadequate polymer, or mixing issues. Remedies include re‑running jar tests, tuning coagulant/polymer dose, and checking mixing intensity. Signs of overdosing include re‑entrant turbidity and floating floc; backing down dose typically helps (studylib.net).
Equipment glitches also derail performance. Verify chemical feed pumps, mixers, and injectors; issues like air in lines, clogged polymer injection, or damaged paddles undermine floc formation. Polymer preparation matters — overly dilute stock can salt‑out — and feed points should be confirmed when any dosing deviation occurs (studylib.net). A simple troubleshooting flow emphasizes checking raw water quality, pH, polymer feed, and injection points before larger changes (studylib.net).
Fragile floc and short settling indicate excessive shear; solutions include reducing mixing intensity or extending flocculation time (pdfcoffee.com). Rising sludge or thick blankets point to removal issues; more frequent sludge withdrawal and clarifier rake repairs typically resolve it (pdfcoffee.com). Hydraulic short‑circuiting calls for inlet baffles and weir checks.
Residual color and natural organic matter (NOM) sometimes persist despite coagulation; optimizing coagulant type/dose can help, but downstream adsorption or oxidation is occasionally needed. Many mills introduce activated carbon as a polishing step when organics are unusually high.
In practice, systematic troubleshooting blends observation and lab retesting. If performance dips, operators re‑run jar tests mimicking full‑scale conditions and consult operations manuals that highlight stepwise checks (pdfcoffee.com; pdfcoffee.com). Consistent records of turbidity, pH, alkalinity, and chemical use help spot trends such as escalating dose needs as raw quality declines.
When the train runs well, the numbers are emphatic. A jar‑test‑optimized plant can take hundreds of NTU down to well under 1 NTU after filtration, with suspended solids (TSS) falling by 90–99% through C/F. One alum jar‑test study reported ~94% turbidity removal at the optimal dose (researchgate.net).
Citations and local variability
Peer‑reviewed studies and manuals underpin these methods and figures, including jar‑test procedures and removal efficiencies (researchgate.net; link.springer.com; researchgate.net; jurnal.ugm.ac.id; studylib.net; pdfcoffee.com). Indonesian water‑quality reports underscore the variability challenge from raw sources (nyjxxb.net). Plants often blend this literature with local jar‑test data to set coagulant and flocculant strategies that stand up to seasonal swings.
