LLDPE line start-up and grade changeover—sequence, purge and first-metre checks—is a multistep operation governed by melt viscosity displacement, screw hold-up, pelletizer thermal uniformity, and analytical verification of the initial pellet bed. On a finishing line using a co-rotating twin-screw extruder with an L/D of 40:1 and a gear pump feeding an underwater pelletizer, barrel zones are preheated to 200–230 °C for a butene-based LLDPE having a melt index of 1.0 g/10 min at 190 °C under 2.16 kg load by ASTM D1238, procedure A, and a density of 0.918 g/cm³ by ASTM D1505. The die plate is held at a minimum of 220 °C before melt first reaches the die holes, and the hopper is inerted with nitrogen to an oxygen concentration below 5 vol%. The line is started at 60–70 % of nominal screw speed, and the screen changer differential pressure must remain below 80 bar before screw speed is increased. Pelletizer water temperature is set at 40–60 °C during start-up because colder water promotes die-face freeze-off and irregular pellet edges.

Upstream reactor transition is often synchronized with downstream finishing. In gas-phase LLDPE production, grade changeover is executed by changing hydrogen-to-ethylene and comonomer-to-ethylene ratios. The resulting powder melt index and density move toward the new target only after 2–4 resin-bed turnovers, and the finishing line can receive powder with a moving melt index. Therefore, extruder start-up and first-metre checks may be performed on a feed stream that is not yet at full target composition. Under these conditions, extruder-feed powder should be sampled every 15–30 min for melt index by ASTM D1238 and density by ASTM D1505 before the extruder is brought to full rate. First-metre pellet checks serve as a downstream confirmation of the reactor transition tail and define the boundary for off-spec segregated storage.

Extruder clean start-up after maintenance is a separate sequence. In a clean system, the screw is initially empty. The barrel is preheated to 200 °C and held for 30 min to stabilize thermal expansion. The first resin introduced is the target grade, not a purge. Because the screw flights are not wetted, the initial torque is high and the screw speed is kept below 30 rpm until the gear pump suction pressure stabilizes between 5 bar and 15 bar. The first pellets often contain air voids and oxidized resin from dead spots around the die plate. These pellets are segregated until bulk density and moisture meet first-metre criteria.

Can a Single Purge Grade Satisfy All LLDPE Transition Paths?

A single purge grade is usually inadequate if the transition crosses a density span greater than 0.008 g/cm³ or a melt-index ratio greater than 10:1. In the change from a 0.5 g/10 min, 0.926 g/cm³ hexene-based LLDPE to a 20 g/10 min, 0.916 g/cm³ butene-based LLDPE, the existing melt pool is first displaced with a fractional-melt-index, high-viscosity polyolefin. That purge is processed at 20–40 rpm on a 75 mm co-rotating twin-screw extruder. The function is mechanical scraping of the screw root and barrel wall while raising wall shear stress in the screw channel. The high-viscosity purge is followed by a low-viscosity purge grade, typically a 20–50 g/10 min LDPE or a broad-molecular-weight linear resin, with a viscosity ratio of at least 20:1 relative to the target grade. The second purge displaces high-viscosity residues from dead zones at the gear pump inlet and screen changer channels. Published purge-volume data for every specific line configuration is limited; in practice, purge volume is adjusted by extruder hold-up multiples. A displacement volume of 6–8 times the combined screw and gear pump volume is used as the minimum estimate for a complete transition. Visual confirmation by colour is not sufficient because the first metres of target grade may contain colour-carrying purge from only the centre of the melt channel while the wall layer remains as a slow-releasing stagnant film.

If the transition is executed in reverse, from a high-melt-index grade to a fractional-melt-index grade, the high-viscosity purge is introduced more slowly. Screw speed is limited by torque and die pressure is monitored because high-viscosity material can exceed the gear pump seal limit. A gear pump on a 75 mm line may have a continuous seal limit of 180 bar. Start-up die pressure is therefore maintained below that limit. The initial purge is not a low-viscosity resin because it can channel through the high-melt-index molten polyolefin and fail to sweep the barrel wall. The sequence remains high-viscosity purge, low-viscosity purge, target grade. The same sequence is not optimal for a clean-line start-up after maintenance, when the screw is initially empty and the first resin introduced acts as its own purge.

Additive masterbatch feed lines are purged before the base powder transition. Residual slip agent, antiblock, or fluoropolymer processing aid from the previous grade can alter pellet surface friction and water separation. The additive feeder is bypassed or switched to additive-free base resin for at least 30–60 s of residence time, or for the time required to discharge two feeder hold-up volumes. Sequential purge alone does not remove adhered additive melt from the die face. A manual die-face wipe or a silicon-free die-face cleaning compound is used when surface friction changes are observed. When the target grade contains a fluoropolymer processing aid, first-metre checks should include melt fracture assessment on a 50 µm cast film because residual purge can delay the processing aid effect and shift surface appearance. Film defects are assessed by ASTM D3351 or equivalent optical scanning.

After the barrel and screen changer are purged, the line enters the most critical period for first-metre pellet formation. Die plate temperature distribution is checked with contact thermocouples. A radial temperature gradient greater than 5 °C across the die plate circumference can produce different pellet lengths from centre and edge holes. Pellet geometry changes before bulk melt index shifts, because hole fill fraction and die swell respond to viscosity differences earlier than the averaged melt index. Operators record die pressure and melt temperature at 1-min intervals until both are stable within ±5 bar and ±2 °C. The first pellet sample is not taken from the cutter water overflow line because water carryover and surface moisture affect bulk density. Instead, the sample is obtained after the centrifugal dryer discharge with the dryer air temperature at 80–100 °C. Pelletizer blade tip speed is maintained at 8–15 m/s, depending on die-hole diameter and melt viscosity. For a 2.4 mm die hole, blade tip speed below 8 m/s often yields elongated or connected pellets, while speed above 15 m/s increases fines from brittle fracture at the die face.

When Melt Index Falls Below 0.5 g/10 min, Torque-Based Purge Staging Becomes Decisive

For fractional-melt LLDPE grades with MI₂.₁₆ below 0.5 g/10 min, start-up is torque-limited rather than output-limited. The screw speed ramp is constrained by extruder drive torque and by the pressure limitation of the gear pump. On a 75 mm twin-screw extruder with a 45 mm gear pump and a 220-hole die plate, die pressure is held below 180 bar during start-up. If the die plate has holes of 2.4 mm diameter, the melt temperature at the die is maintained at 230–250 °C to prevent hole freeze-off, while barrel zones are set 10–20 °C lower to limit viscosity reduction along the screw. Specific energy input for LLDPE during start-up is typically 0.18–0.22 kWh/kg. Sustained operation above 0.24 kWh/kg can generate localized hotspots in kneading blocks and produce gel particles in the first pellet fraction. Target fractional-melt grade is introduced only after low-viscosity purge has reduced screen changer differential pressure below 60 bar. Cooling to the standard die set point is delayed until first-metre checks confirm acceptable gel count and pellet geometry, because early cooling increases die pressure and may exceed the seal limit. Melt-flow ratio I₂₁.₆/I₂.₁₆ by ASTM D1238 is used to check for crosslinking or chain scission. A shift of more than 10 % from the reference is an off-spec criterion.

Pelletizer water temperature is raised by 5–10 °C for fractional-melt LLDPE because the pellet surface is more sensitive to chill marks. Water flow across the die face is adjusted so that the pelletizer inlet water temperature does not fall below 45 °C. The first metres of fractional-melt grades may show 1–3 % fines under stable cutting. If fines exceed 6 %, blade speed is trimmed before further analytical testing begins. The first-metre sample is segregated until the melt-flow ratio and pellet geometry return to reference.

Sieve Analysis, Bulk Density and Pellet Geometry in First-Metre Sampling

First-metre checks are not limited to melt index and density. The first sample is split from the segregated pellet fraction and screened according to ASTM D1921. A typical LLDPE pellet target contains less than 6 % by mass below 2.36 mm and less than 4 % by mass above 4.75 mm. Bulk density measured by ASTM D1895 method A should fall within 0.48–0.55 g/cm³ for most LLDPE pellet grades. A value below 0.48 g/cm³ suggests void-containing pellets, connected pellets, or excessive fines. Moisture content is measured by ISO 15512; the first-metre release limit is below 0.05 % by mass for silo transfer. Pellet aspect ratio, measured by image analysis, is held between 0.75 and 1.25. A shift in aspect ratio during the first metre correlates with die-face blade alignment rather than resin grade, and it is corrected by adjusting blade-to-die contact gap or blade speed before further analytical testing.

First-metre pellet analysis is used as a time-resolved process diagnostic, not as a final release sample. If the pellet size distribution exceeds the fines threshold, the line is held at reduced screw speed and pelletizer blade speed is trimmed in 0.5 m/s increments. If moisture exceeds 0.05 %, dryer air temperature is raised by 10 °C and pellet residence time is increased. If melt index of the first metre is outside ±10 % of target, the line returns to low-viscosity purge for one additional hold-up volume. These adjustments are performed on the segregated first-metre fraction. Steady-state silo samples collected after 4 h of operation must meet release specifications, but first-metre data determine whether the transition can proceed without contaminating downstream storage. On a multi-grade LLDPE finishing line, a die-plate temperature drop of 8 °C during the first metre produced a fines increase to 9 % and required 30 min of off-spec segregation before pellet geometry returned below the 6 % fines threshold. That behaviour is characteristic of edge-hole freeze-off in a 220-hole die plate.

Operational boundaries for LLDPE start-up include minimum pelletizer water temperature of 40 °C, maximum die pressure of 180 bar for many gear-pump-assisted lines, and hopper oxygen below 5 vol%. Transition between a narrow-molecular-weight metallocene LLDPE and a Ziegler-Natta LLDPE of identical melt index and density may require longer purge because differences in die swell and melt strength persist at shear rates below 100 s⁻¹. The purge sequence cannot remove adhered processing aid from the die face; a manual die-face wipe or a silicon-free cleaning compound is required when pellet surface friction changes. Moisture-sensitive additive masterbatches must be purged separately before the final grade is introduced. Published data for the exact transition time of every formulation is limited; the sequence is therefore tuned with hold-up multiples and first-metre analytical feedback rather than by fixed time alone.