Plastic Recycling Industry

Three streams feeding the same extruder — clean post-industrial PE scrap, washed agricultural film and mixed packaging bales — carry completely different moisture profiles. Clean scrap usually stays below 0.05%, while agglomerated agricultural film wanders between 0.4 and 1.2% even after the mechanical dryer. Applying one let-down to all three means waste on one and bubbles on the other two.

Take the sample at the extruder throat, not at the silo. Between the dryer outlet and the throat, the conveying route puts measurable water back onto the flake surface whenever ambient relative humidity exceeds 65%; on cold morning shifts, condensation forms on the silo wall and the first hours of pellet come out visibly worse.

Sample each stream over at least five shifts at different times of day, and record the result as a range rather than a single average. The dosing decision is made on the upper percentile, not the mean; a line tuned to the average produces bubbles a third of the month, and that loss never appears in the cost sheet.

A halogen moisture balance is fast but at 105–130 °C it sees only water evaporating from the surface. Water held in agglomerate pores and on filler surfaces does not come out at that temperature; meanwhile volatile organic residue and low-molecular-weight oils join the mass loss and push the reading up instead. Because the two errors work in opposite directions, the result becomes unpredictable rather than merely low.

The reference method is coulometric Karl Fischer titration with a 160–180 °C oven headspace: it measures water specifically and is not confounded by volatiles. On the same sample it is common to see loss on drying at 0.15% and Karl Fischer at 0.40–0.50%. Running both once in parallel and deriving a plant-specific correction factor is the practical way to keep daily control on the fast method.

Repeat the measurement on product changeover, on supplier change and at season transitions. The difference between summer and winter can double residual moisture on the same material through the same wash line — enough on its own to shift the whole dosing window.

The calculation is one line: water to be bound (kg/t) × 3.11 = theoretical CaO (kg/t). Because dispersion and diffusion in the melt are limited, real efficiency is 55–70%, so multiply the theoretical figure by 1.5–1.8. For 0.10% residual moisture: 1 kg water × 3.11 × 1.6 ≈ 5 kg CaO per tonne, i.e. 0.5% active CaO. With a 50% CaO masterbatch that is a 1.0% let-down.

When choosing the masterbatch, check carrier compatibility with the base resin: LDPE/LLDPE carrier on a PE line, PP or a very-low-density copolymer carrier on a PP line. An incompatible carrier spoils dispersion and CaO agglomerates show up as gels in film. The carrier melt flow index should sit close to that of the base resin; a large gap makes the masterbatch behave as a separate phase in the melt.

Meter the dose gravimetrically. With volumetric dosing, a 5–8% change in the bulk density of the masterbatch granules feeds straight through into the dose — for a moisture scavenger that deviation is large enough to void the stoichiometry you just calculated. Recalibrate the feeder whenever the masterbatch batch changes.

Fineness is the intersection of two independent constraints. The first is kinetic: hydration must reach the core of the particle within a melt residence time of 30–120 seconds, and d50 of 3–6 µm with a BET of 1.5–4 m²/g fits that window. The second is optical: in blown film 20–40 µm thick, a particle above 25 µm creates a lens defect and the start of a pinhole. The two constraints point to the same range not by coincidence but by physics.

Hard-burnt, low-porosity lime may carry the same available CaO on its certificate and still be too slow in the melt. Lime that cannot finish reacting stays in the pellet as free CaO, draws ambient moisture in storage and swells the pellet — you end up adding a new defect to material you believed you had dried. Specify soft-burnt, porous lime and require reactivity on the certificate.

Where colour matters, look at Fe₂O₃ and total impurities; in clear and white film a few hundred ppm of iron becomes a visible yellowish cast. The same purity logic governs every field where lime works as a chemical reactant — the specification reasoning on the chemical and pharmaceutical industry side rests on exactly the same cause-and-effect chain as this page.

The scavenger masterbatch works after the vent, not instead of it. The vacuum vent removes steam released in the melting zone; CaO binds the water that opens out of an agglomerate pore downstream of the vent, where there is no longer any way to expel it. Treating the two as alternatives is the most expensive mistake on site: switching off the vacuum and raising the let-down triples chemical cost and degrades mechanical properties on top.

Typical melt temperature is 180–230 °C for PE and 200–240 °C for PP. Lowering temperature to fight bubbles usually backfires: melt viscosity rises, dispersion deteriorates, CaO agglomerates grow and gel count climbs. Bubbles are solved with water content, dispersion and venting efficiency — not with the heater setpoint.

Target −0.6 to −0.9 bar at the vent and check the vacuum line for blockage weekly. Melt escaping at the vent port usually comes from overfeeding or overfilling and quietly kills the vacuum; the first symptom is a bubble, and the lime dose gets blamed again. Track screen-changer pressure rise rate on the same chart.

Used at 10–40%, natural calcium carbonate filler raises stiffness, increases thermal conductivity so the cooling line runs faster, and lowers unit cost. But the calcite surface is hygroscopic; uncoated or openly stored filler carries 0.1–0.2% water into the compound on its own. A filler coated with 1.0–1.2% stearic acid, d50 of 1.5–3 µm and a top cut below 12–15 µm, markedly improves both dispersion and moisture pickup.

Filler particle size distribution governs more than moisture — it sets surface and optical behaviour too. For another field where the same cause and effect plays out on the filler–opacity axis, see the paper and cellulose industry; the particle-size-to-surface relationship there rests on the same basis as the gel and lens defect logic here.

Scavenger masterbatch arrives in aluminium-foil-laminated moisture barrier bags and starts losing activity the moment it is opened: at 60% relative humidity, measurable loss begins within 6–8 hours in an open bag. Open only what a shift needs and reseal the rest airtight. Binding ambient humidity in the silo, warehouse and container with a moisture and gas absorbent directly extends the masterbatch's usable life.

Do not hang verification on a single indicator. Vacuoles in the pellet cross-section, gel and lens counts per square metre of film, pinhole count, haze, melt pressure fluctuation and screen-changer pressure rise rate are read together. When the dose is right, melt pressure settles, die drool falls and screen life extends — these tell you the dose is correct earlier than any bubble count.

Overdosing looks different and is easily attributed to something else: white specks in the pellet, rising gel count in film, falling impact strength and elongation at break, and inconsistent printing and heat sealing. Unreacted CaO left in the pellet draws moisture in storage and cracks the pellet open. With a moisture scavenger, the rule that “a little extra never hurts” does not hold.

Do not set the recipe once and forget it: the dose is recalculated as feedstock source, season and wash-line performance change. Deriving a specification and dosing window for your own moisture profile means doing it on your own Karl Fischer data rather than a masterbatch supplier's generic recommendation; to work through that together, get in touch with us.