Starch & Fiber Recovery

PAM for Starch Recovery in Paper Mills

Mills lose starch at the size press, in the coating kitchen, through broke handling and in wet-end carryover. Each of those losses shows up twice on the cost sheet: once as raw material bought and not used, and again as chemical oxygen demand arriving at the effluent plant. Starch is highly biodegradable, so a mill that loses starch is simultaneously paying for size and overloading its own biological treatment.

Polyacrylamide is part of the answer, but only part, and the boundary matters enough that we put it before the specifications rather than after them. We supply anionic and cationic PAM from our own plant in Xinxiang, Henan, with 100,000 tons/year of PAM capacity, and we would rather you dose the right stream than dose the wrong stream generously.

What Polymer Can and Cannot Recover

Flocculation only acts on material that has a surface. Suspended and colloidal solids — ungelatinised or retrograded starch granules, starch-coated fines, filler-starch agglomerates, coating colour carryover — have surfaces that polymer chains can adsorb onto and bridge. Those you can capture, and capture efficiently.

Truly dissolved starch does not. Cooked, solubilised amylose and amylopectin are molecularly dispersed in the water phase, and no dose of PAM will pull them out, because there is no particle to aggregate. A supplier promising to recover 70-85% of the starch in your effluent with 2-4 ppm of anionic PAM has either not distinguished the two fractions or is quoting a suspended-solids number and calling it starch.

So the first task on any starch-bearing stream is not dose selection — it is splitting the load. Filter a sample through a 0.45 micron membrane and run chemical oxygen demand on the raw sample and the filtrate. The difference is the particulate fraction that polymer can act on; the filtrate COD is the dissolved fraction that needs biological treatment, evaporation or membrane concentration instead. Mills are frequently surprised by the split, and it changes what the right project is.

• High particulate share — a saveall or DAF unit with correctly chosen PAM is the cheap win.
• High dissolved share — polymer will clarify the water cosmetically while COD barely moves. The project is anaerobic pretreatment or loop closure, and polymer is a polishing step at the end of it.

The split also moves over time, and not in your favour. Amylase enzymes used deliberately to thin cooked starch do not stop working when the starch reaches the sewer, and warm nutrient-rich process water grows the organisms that produce more of them. Starch that was particulate when it left the size press can be substantially solubilised a few hours later in a buffer tank. That has a design implication worth stating plainly: capture close to the source, because residence time converts recoverable solids into unrecoverable dissolved load. A saveall taking a short, warm, segregated stream will outperform the same unit sited at the end of a long combined sewer, on identical polymer.

Where the Starch Is Actually Going

Two of those five rows are not chemistry problems. Segregating a strong, clean size press stream for reuse inside the loop recovers starch at full value, which no downstream treatment can match. Buffering a slug of cooker washout so it feeds the biological plant over hours instead of minutes prevents the shock loading that causes most starch-related permit excursions. Do both before costing a polymer programme, because they are cheaper and they shrink the problem the polymer has to solve. Reducing wet-end carryover is a retention question rather than a recovery one, and we treat it separately under paper making retention and drainage.

Grade Selection by Stream

The units in the last row are not comparable to the rows above it. Clarification and flotation doses are grams of polymer per cubic metre of water — parts per million. Dewatering dose is kilograms per tonne of dry solids in the sludge, which on a thin feed can be a thousand times more polymer per unit volume. Quoting a dewatering figure as though it were a clarification figure is how polymer budgets end up wrong by an order of magnitude; the arithmetic for converting between the two bases is set out in our dosage calculation guide. Full specifications for these grades are on the cationic polyacrylamide and anionic polyacrylamide pages, and the reasoning behind picking one over the other is in anionic vs cationic polyacrylamide.

Matching Polymer to the Separation Equipment

The same water needs a different polymer depending on how you intend to separate the floc, because each device rewards a different floc property. Choosing polymer before choosing equipment, or copying a dose from a mill running different hardware, is a common reason a trial underperforms on paper that looked identical.

• Dissolved air flotation wants light, air-attaching flocs of moderate size. Excessive molecular weight builds dense flocs that sink instead of floating, so a mid-range CPAM outperforms the highest-MW grade in the catalogue. This is the most common duty on coating and starch washwater, and the one where over-specifying molecular weight actively hurts.
• Gravity clarifiers and lamella settlers want the opposite — dense, fast-settling flocs. Higher molecular weight anionic PAM behind a coagulant is the right build, and settling velocity rather than clarity is the number to optimise on the bench.
• Saveall disc and drum filters want flocs strong enough to survive the approach to the filter medium and porous enough not to blind it. Overdosing here shows up as a rising differential pressure and falling throughput rather than as poor clarity, which makes it easy to misdiagnose as a mechanical fault.
• Belt presses and screw presses on recovered sludge want high charge density cationic PAM at the dewatering dose basis, sheared enough to release water but not so much that the cake smears. This is a separate chemical from the clarification duty upstream and should be costed separately.

A practical consequence: if you are running a DAF and a press on the same recovered stream, expect to buy two grades rather than one. Mills that try to serve both duties with a single product usually end up over-dosing the flotation stage to make the press behave, which wastes polymer at the point where water volume is largest. Charge density is the property that separates the two duties, and how to read it on a specification sheet is covered in our charge density explainer.

The COD Load Is Usually the Real Money

Starch carries roughly 1 kg of chemical oxygen demand per kilogram of starch, so losses translate almost directly into treatment load. That has three cost consequences that are easier to quantify than the value of recovered starch itself.

• Aeration energy. Aerobic treatment of a readily biodegradable load consumes blower power in proportion to the COD removed. Removing particulate load ahead of aeration removes that power draw.
• Sludge yield. Biodegradable COD converts to biomass. Every kilogram treated biologically becomes sludge you dewater and dispose of, whereas a kilogram removed as a floc in a DAF unit is already a concentrated solid.
• Surcharge or permit headroom. Mills discharging to a municipal works pay by load. Mills treating on site pay in permit risk when a starch slug arrives faster than the biology can absorb it.

This is why we frame starch projects as load reduction rather than raw material recovery. The load saving is measurable next month on an energy bill and a sludge haulage invoice. The raw material saving depends on whether the recovered material is fit to reuse, which is a harder question.

Whether Recovered Starch Can Go Back Into Size

Usually not into surface size, and it is better to say so up front. Recovered starch solids come back contaminated with fibre fines, filler, coating pigment and biological growth, and the starch itself has been partly degraded by shear, heat and enzymatic attack in the loop. Feeding that into a size press risks runnability problems and visible surface defects, and the size press is the wrong place to take a quality gamble.

There are three realistic destinations. Recovered solids can go back into the furnish as filler-and-fines mass on grades that tolerate it, which recovers the fibre and filler value even if the starch value is discounted. A clean, segregated stream captured close to the source can sometimes return to wet-end starch duty, where the specification is far more forgiving than surface sizing. Or the solids are dewatered and disposed of, in which case the entire benefit is the COD reduction described above — which is often still worth doing.

Decide which of the three you are aiming at before selecting equipment, because it changes the polymer choice. Recovery for furnish reuse wants tight, clean flocs and no residual cationic charge that would upset wet-end chemistry. Recovery purely for disposal wants whatever dewaters cheapest.

Worked Example: Sizing the Prize Before You Spend

Modelled from process arithmetic, not a delivered project. Take a mill losing 500 kg per day of starch solids to effluent. The membrane-filtration split comes back 40% particulate, 60% dissolved — a typical shape when size press losses dominate.

The particulate 200 kg/day is what a DAF unit with correctly chosen CPAM can address, and at a realistic capture efficiency of 80-90% on the particulate fraction that is 160-180 kg/day of solids removed from the biological load. Note what happened to the headline: the honest recovery figure against total starch loss is 32-36%, not the 70-85% that a specification sheet reading only suspended solids would suggest. The remaining 300 kg/day of dissolved starch stays in the water and still has to be treated biologically.

The value of those 160-180 kg/day sits in three places — aeration power not consumed, biological sludge not generated and dewatered, and load surcharge or permit headroom not spent. Put your own numbers against each: your electricity tariff, your sludge disposal cost per wet tonne, your discharge tariff. Polymer cost for a DAF at 1-4 g/m³ is usually a small fraction of that total, which is why these projects tend to pay back quickly when the particulate share is high — and why they disappoint when it is low. Run the split test first, then the arithmetic, then the enquiry.

Dosing Practice on Starch-Bearing Streams

1. Run the filtered-COD split before anything else. It tells you whether polymer is the project or a footnote to the project, and it takes one afternoon.
2. Jar test on the real stream, at the real temperature. Starch water is often 40-55°C. Polymer behaviour and floc strength both change with temperature, so a bench test at ambient will mislead you. The procedure, including the shear step operators usually skip, is in our jar test procedure.
3. Coagulate before you flocculate. High-MW polymer on raw, high-charge-demand water is expensive and unstable. A cheap fixative or inorganic coagulant first does the charge work; the PAM then only has to bridge. Our polyDADMAC guide covers the fixative side.
4. Make down fresh and use within 24 hours. Starch water is nutrient-rich and warm, so microbial degradation of both the polymer solution and any carryover is faster than in a municipal plant. Make down with chlorine-free water at 0.2-0.5% and post-dilute at the injection point.
5. Watch for the overdose turn. Past the optimum, charge reversal redisperses flocs and turbidity climbs. If more polymer is making the water worse, the answer is less polymer.
6. Hold your supplier to batch consistency. We test every batch for molecular weight to a ±0.5M tolerance and retain samples for 24 months, so when performance shifts we can compare your drum against the retained sample and establish whether the polymer or the stream changed. Chain length is the property that drifts most in trading-house material, and the reason it matters is set out in our molecular weight guide.

Standards and Test Methods We Work To

The measurements that decide a starch recovery project are effluent measurements, and they are defined by published methods rather than supplier claims. Chemical oxygen demand, total and dissolved solids and turbidity follow the methods in Standard Methods for the Examination of Water and Wastewater, published jointly by the American Public Health Association, the American Water Works Association and the Water Environment Federation. Paper-side testing follows TAPPI and ISO methods. Zellcheming test methods are widely referenced by European mills, and CEPI publishes sector data useful for benchmarking specific water and load figures against comparable mills.

For products used where treated water re-enters a food-contact grade loop, residual acrylamide monomer limits are set by the applicable national food-contact regulation in your market. We supply a certificate of analysis showing monomer content for each batch so compliance can be demonstrated rather than asserted. We do not quote method or document numbers on this page, because designations get revised and a stale number stated with false precision is worse than a named body you can verify yourself. Ask us against your specific test and we will confirm the current designation in writing.

Frequently Asked Questions

Can PAM recover dissolved starch from my effluent?

No. Flocculation acts on particles with a surface. Dissolved amylose and amylopectin are molecularly dispersed, so there is nothing to aggregate and no dose will change that. PAM recovers the particulate fraction — granules, starch-coated fines, pigment-binder agglomerates. Split the load with a 0.45 micron filtered COD test before sizing any polymer programme.

A supplier quoted 70-85% starch recovery. Is that credible?

Only if it refers to the particulate fraction. Against total starch loss it is usually not achievable, because the dissolved share is commonly half or more of the load and polymer cannot touch it. Ask which fraction the percentage is measured against, and ask for the filtered COD split that supports it.

Cationic or anionic on starch-bearing water?

Cationic where the target is negatively charged suspended solids such as coating pigment and fines, which is most saveall and flotation duty. Anionic high molecular weight PAM works behind an inorganic coagulant or a cationic fixative, where charge neutralisation is already done and the polymer only needs to bridge. On raw high-charge-demand water, anionic alone underperforms.

Can recovered starch go back into surface size?

Generally not. Recovered solids carry fibre fines, filler, pigment and biological growth, and the starch has been partly degraded by heat, shear and enzymatic attack. Realistic destinations are furnish reuse as fines-and-filler mass, wet-end starch duty from a clean segregated stream, or dewatering and disposal with the benefit taken as COD reduction.

Does temperature affect polymer selection here?

Yes, and it is routinely overlooked. Starch streams often run 40-55°C, which accelerates polymer hydration but also weakens flocs and speeds chain degradation in the make-down tank. Jar test at process temperature, not at ambient, and shorten solution hold time compared with a cold municipal application.