Building a PLA twin screw compounding factory is a civil engineering project wearing a machinery costume. Investors who approach the total construction and equipment budget for PLA twin screw compounding factory projects by first asking “how much is the extruder” almost always mis-plan the site, because the extruder is only one of roughly a dozen budget families that must be sized together. Floor loading, ceiling height, transformer capacity, cooling water circulation, dehumidifying drying, pellet conveying, laboratory instrumentation, spare parts, commissioning labor and working capital all scale with the capacity tier you choose on day one. Change the tier and every line moves.
This guide is written by the engineering team at Kerke, a Wanplas factory that has spent 12 or more years designing and building co-rotating parallel twin-screw compounding systems. From a 19,997 square meter manufacturing base with more than 100 team members, Kerke has placed over 2,000 machines into production across 70 or more countries, and biodegradable compounding — PLA, PBAT, starch-filled and mineral-filled systems — has become one of the fastest growing segments of that installed base. The observations below come from configuring, testing and commissioning those lines, not from a spreadsheet template.
Because currency figures age badly, vary by region, and mislead readers comparing quotations across continents, this article deliberately uses no monetary amounts at all. Instead the entire plant investment is normalized to 100 index points. Each budget family is expressed as a range of index points, so you can see immediately which decisions actually move the total and which are rounding noise. Relative grades — Low, Medium, High, Very High, Premium — describe intensity. Payback is discussed in years and months, capacity utilization in percent, and quality performance in physical units such as parts per million moisture or molecular weight retention. Apply your own local rates to the index structure and you get a live budget that stays valid in 2026 and beyond.
What follows is a complete capex decomposition: capacity tier definition, the index point allocation table, civil works, utilities, the extruder and its ancillaries, drying, pelletizing route economics, downstream handling, the quality laboratory, PLA-specific process engineering, real Kerke KTE series configurations with full specification tables, a requirement-to-model selection table, manpower and energy intensity, a month-by-month construction schedule, a risk ranking of overrun candidates, and a phased-build strategy that protects the option to double capacity later.
Defining the Project: Capacity Tier Drives Every Budget Line
Nothing in a compounding plant budget can be estimated until the annual capacity tier is fixed, because capacity determines screw diameter, screw diameter determines motor rating, motor rating determines transformer size, and transformer size determines the entire electrical civil package. A PLA compounding project is therefore defined top-down: annual tonnage, then operating pattern, then hourly throughput, then machine size, then everything else.
Four tiers cover most greenfield biodegradable compounding investments. The 500 ton per year tier is a market-entry or captive-supply plant, often built by a converter who wants formulation control over the resin feeding its own bag or film lines. The 1,000 ton tier is the classic first commercial line. The 2,000 ton tier is where a plant becomes a genuine merchant compounder with multiple recipes and customers. The 5,000 ton tier is an industrial installation with dedicated silos, bulk logistics and, usually, two or more extrusion lines rather than one very large machine.
The conversion from annual tonnage to hourly throughput is where optimistic plans die. A realistic first-year plan should assume 300 operating days, two or three shifts, an overall equipment effectiveness in the 70 to 82 percent band, and a first-pass yield of 95 to 98 percent once the recipe is stable. Recipe changeovers on a multi-product PLA line consume real hours: purge material, screen change, feeder recalibration and the first-article laboratory check together typically take 90 to 180 minutes depending on how far apart the two recipes sit.
Capacity Tier Definition and Required Line Throughput
| Parameter | Tier A — 500 t/y | Tier B — 1,000 t/y | Tier C — 2,000 t/y | Tier D — 5,000 t/y |
|---|---|---|---|---|
| Operating pattern | 1–2 shifts, 300 days | 2 shifts, 300 days | 3 shifts, 300 days | 3 shifts, 330 days |
| Scheduled hours per year | 3,600–4,800 | 4,800 | 7,200 | 7,900 |
| Assumed effectiveness | 70–75% | 75–80% | 78–82% | 80–85% |
| Required rated output | 180–220 kg/h | 280–330 kg/h | 380–450 kg/h | 800–950 kg/h |
| Typical line count | 1 | 1 | 1 (2 if multi-recipe) | 2 |
| Building footprint guide | 900–1,400 m² | 1,400–2,000 m² | 2,200–3,000 m² | 4,000–6,000 m² |
| Capital intensity per annual ton | Very High (index 100 base) | High (0.62–0.70 of Tier A) | Medium (0.44–0.52 of Tier A) | Low (0.30–0.38 of Tier A) |
The last row is the single most important line in this article. Capital intensity per annual ton of capacity — the total investment index divided by nameplate tonnage — falls steeply as the tier rises. Moving from Tier A to Tier D cuts the index points required per annual ton by roughly two thirds. This is why undersized “test the water” plants so often disappoint: they carry nearly the full civil, utility and laboratory burden of a real plant while producing a fraction of the tonnage over which to spread it. If your market plan can justify Tier B or Tier C, the per-ton economics improve dramatically without a proportional increase in absolute investment.
The Index Point Method: Reading a Plant Budget Without Currency
The index point method normalizes total project investment to 100 points and distributes those points across budget families. It is the only comparison method that survives translation across countries, currencies and years, and it is how experienced project engineers sanity-check a quotation package before any negotiation begins. If a supplier’s proposal implies that the pelletizing and cooling package is worth 2 index points, something has been left out of the scope.
Below is the master allocation table for a PLA and PBAT compounding plant, shown for the four capacity tiers. The point totals for each tier sum to 100 within rounding; the ranges reflect the genuine spread between a lean domestic-market build and a fully instrumented export-grade plant.
Master Budget Allocation — 100 Index Point Model
| Budget family | Tier A 500 t/y | Tier B 1,000 t/y | Tier C 2,000 t/y | Tier D 5,000 t/y | Scaling behavior |
|---|---|---|---|---|---|
| Land, civil works, building envelope | 22–28 | 20–26 | 18–24 | 16–22 | Sub-linear; site-dependent |
| Utilities and site services | 10–14 | 10–14 | 11–15 | 12–16 | Rises with connected load |
| Twin-screw extruder main machine | 16–20 | 17–21 | 18–23 | 19–24 | Near-linear with size class |
| Feeding and dosing (loss-in-weight) | 4–6 | 4–6 | 5–7 | 5–8 | Driven by recipe complexity |
| Devolatilization, melt pump, screen changer | 3–5 | 3–5 | 4–6 | 4–6 | Near-linear |
| Pelletizing and pellet water circuit | 4–6 | 4–7 | 5–7 | 5–8 | Route-dependent |
| Drying and dehumidification | 4–6 | 4–6 | 4–7 | 5–7 | Mandatory for PLA |
| Downstream: screening, conveying, packing | 3–5 | 4–6 | 5–7 | 6–9 | Rises with automation level |
| Quality laboratory and instruments | 3–5 | 3–4 | 2–4 | 2–3 | Nearly fixed; dilutes with scale |
| First-fill spare parts | 2–3 | 2–3 | 2–4 | 3–4 | Proportional to machine value |
| Installation, commissioning, training | 3–5 | 3–5 | 3–5 | 3–5 | Roughly constant share |
| Trial material loss and start-up scrap | 2–4 | 2–4 | 2–4 | 2–4 | Recipe-count driven |
| Certification and compostability testing | 2–4 | 2–3 | 1–3 | 1–2 | Fixed per recipe family |
| Working capital at start-up | 8–12 | 9–13 | 10–14 | 11–16 | Linear with throughput |
Reading the table: the complete extrusion island — main extruder, feeders, devolatilization train, melt pump, screen changer and pelletizer — occupies roughly 28 to 35 index points. Buildings and utilities together occupy 28 to 40. Everything else, including the laboratory, spares, commissioning and working capital, occupies the remaining 25 to 35. A budget that shows the machine at more than half of total investment has almost certainly under-scoped the building or the utilities.
Land, Civil Works and the Building Envelope
Civil works consume 16 to 28 index points and are the single most commonly underestimated family in a compounding plant budget. A twin-screw compounding hall is not a generic warehouse: it needs vertical space for gravimetric feeder towers and drying hoppers, a floor slab that will not crack under a heavy extruder base, dedicated fire compartments for combustible resin storage, and cable and pipe routing that does not obstruct screw pulling.
Clear internal height is the first hard requirement. Loss-in-weight feeders must sit above the extruder feed throat, day hoppers sit above the feeders, and vacuum loaders sit above the hoppers. Stack that column and 7 meters of clear height under the crane hook is a practical minimum; 8 meters is comfortable and gives room for a larger drying hopper when you later move to a bigger tier. Retro-fitting height is impossible, so this is a decision that must be right the first time.
The second requirement is floor loading. Under the extruder base frame and the gearbox, design for 8 to 12 tons per square meter, with an isolated foundation pad where local soil conditions are poor. The pelletizer, cooling tank and vibrating screen also concentrate load and vibration. A grouted, level base plate with anchor bolts cast in place is far cheaper than chasing alignment problems in the gearbox coupling for the next decade.
Civil Works Sub-Allocation Within the Civil Family
| Civil sub-item | Share of civil family | Design requirement | Consequence if under-specified |
|---|---|---|---|
| Steel structure and roofing | 30–38% | Clear height 7–8 m; crane beam 5–10 t | Feeder tower will not fit; no screw pulling clearance |
| Floor slab and machine foundations | 18–24% | 8–12 t/m² local loading, flatness class for pellet trolleys | Slab cracking, gearbox misalignment, vibration transfer |
| Resin warehouse and fire compartmentation | 12–18% | Separated storage bay, sprinkler coverage, smoke venting | Insurance refusal, occupancy permit delay |
| Dust-classified zones and static control | 6–10% | Explosion-relief design where mineral powder is handled | Powder ignition risk; rework of the filler dosing bay |
| Utility rooms, transformer yard, pump house | 10–14% | Separate rooms, ventilation, noise attenuation | Overheating switchgear, noise complaints |
| Laboratory, office, welfare, changing rooms | 8–12% | Temperature and humidity controlled lab room | Unstable moisture and MFR test results |
| Yard, roads, loading dock, drainage | 6–10% | Container turning radius, dock leveler, storm drainage | Slow truck turnaround, wet resin bags |
Two civil decisions deserve special attention in a PLA plant. First, the resin warehouse must be a genuinely dry environment, because PLA pellets absorb atmospheric moisture through the bag over time and every point of absorbed moisture must later be removed by the dryer at an energy penalty. Second, if the recipe portfolio includes calcium carbonate, talc or starch, the powder handling bay should be designed as a classified dust zone from the start. Adding explosion relief panels, bonded and grounded equipment and a properly rated dust collector after the building is finished costs several index points more than designing them in.
Utilities: Power, Cooling Water, Compressed Air and Air Treatment
Utilities absorb 10 to 16 index points and act as a hard gate on the schedule: a plant cannot commission until the transformer is energized, and grid connection lead times are frequently the longest single item in the whole project. Utility sizing must be completed at the concept design stage, before civil drawings are frozen, because transformer yard, pump house and cooling tower positions all depend on it.
Electrical capacity is calculated from connected load with a demand factor, never from the sum of nameplate ratings. On a compounding line the main drive motor rarely runs above 65 to 80 percent of rated torque in steady state, barrel heaters cycle at a low duty once the line is at temperature, and the vacuum pumps, feeders and conveying blowers are small but continuous. A practical rule is to take total connected load, apply a demand factor of 0.65 to 0.75, add 20 percent headroom for a future second line, and specify the transformer at the next standard size up. Under-sizing the transformer to save a few index points is the classic false economy, because upgrading later means a second outage, new cabling and a new utility application.
Cooling water is the second gate. A compounding line rejects heat at three points: the gearbox oil cooler, the barrel cooling channels, and the pellet water circuit. These three duties want different temperatures, which is why a single-loop cooling design almost always disappoints. Gearbox and barrel cooling can share a tower water loop at 28 to 32 degrees Celsius, but the pellet water circuit for PLA benefits from chilled water at 12 to 20 degrees Celsius because faster quenching of a slow-crystallizing polymer reduces pellet agglomeration in the dewatering stage.
Utility Sizing Guide by Capacity Tier
| Utility | Tier A 500 t/y | Tier B 1,000 t/y | Tier C 2,000 t/y | Tier D 5,000 t/y |
|---|---|---|---|---|
| Transformer capacity (with headroom) | 250–315 kVA | 400–500 kVA | 630–800 kVA | 1,250–1,600 kVA |
| Tower cooling water flow | 20–30 m³/h | 35–50 m³/h | 55–80 m³/h | 110–160 m³/h |
| Chilled water duty (pellet circuit) | Low | Medium | Medium–High | High |
| Compressed air (7–8 bar, dried) | 1.0–1.6 Nm³/min | 1.6–2.4 Nm³/min | 2.5–3.6 Nm³/min | 5.0–7.5 Nm³/min |
| Dust extraction airflow | 3,000–5,000 m³/h | 5,000–8,000 m³/h | 8,000–12,000 m³/h | 15,000–24,000 m³/h |
| Vent gas treatment need | Low–Medium | Medium | Medium | Medium–High |
| Utility share of total index points | 10–14 | 10–14 | 11–15 | 12–16 |
Air treatment deserves its own paragraph because PLA compounding is not an odorless process. Vacuum devolatilization pulls residual monomer, oligomer, moisture and additive volatiles out of the melt; the vent stream must be condensed and, depending on local emission rules, passed through an activated carbon or catalytic treatment stage. Mineral filler handling generates fine dust that must be captured at source with a cartridge filter unit. Budget for both from the start: retrofitting an emission treatment train after an inspection finding is far more disruptive than building it in, and permit conditions in most jurisdictions in 2026 are stricter than they were even a few years ago.
Steam or thermal oil is optional. Most PLA compounding lines heat the barrel electrically with cast aluminum or ceramic band heaters, which is simpler, faster to control and easier to maintain. Thermal oil heating becomes attractive only where very tight temperature uniformity along the barrel is required, or where a die and melt pump assembly must be held at a precisely uniform temperature. For a first-time PLA plant, electric heating with proportional-integral-derivative zone control and forced-air barrel cooling is the recommended baseline.
The Compounding Core: Twin-Screw Extruder Selection for PLA
The co-rotating parallel twin-screw extruder absorbs 16 to 24 index points and defines the technical ceiling of the entire plant. Every recipe you will ever be able to run, every throughput you will ever be able to reach, and every quality level you will ever be able to certify is bounded by choices made in this specification. For PLA and PLA blends, three parameters matter more than all others: specific torque, L/D ratio, and barrel segmentation.
Specific torque, expressed as T/A³ in newton meters per cubic centimeter, describes how much twisting force the machine can transmit per unit of screw volume. Standard-torque machines sit around 5 to 7 N·m/cm³. High-torque machines used for demanding compounding duty sit in the 8 to 12 N·m/cm³ band. PLA compounding needs the high-torque class for a specific reason: the polymer must be processed in a narrow temperature window, so you cannot simply raise melt temperature to reduce viscosity and unload the drive. Torque capability is what lets you keep the melt cool while still dispersing filler and nucleating agent properly.
L/D ratio determines how many process operations fit along the screw. A PLA and PBAT compound with mineral filler and a chain extender needs, in sequence: feeding, melting, first mixing, side feeding of filler, second dispersive mixing, liquid injection, distributive homogenizing, vacuum devolatilization, and pressure build-up. Fitting all nine zones comfortably requires L/D 40 to 44 as a minimum, and 44 to 48 where reactive extrusion or heavy devolatilization duty is present. Choosing a short 32 to 36 L/D machine to save a few index points permanently restricts the recipe portfolio.
Barrel segmentation and screw modularity are the third pillar. Kerke builds barrels as bolted segments — closed, side-feed, vent, and combination types — with nitrided or bimetallic liners, and builds screws from splined shafts with interchangeable conveying elements, kneading blocks and mixing elements. Kerke’s computer-aided screw assembly design uses kneading co-type geometry with strong self-cleaning behavior and good interchangeability, which means the same machine can be reconfigured for a different recipe family in a shift rather than replaced. That modularity is what protects the investment when the product mix inevitably shifts.
Wear Protection: Where to Spend and Where Not To
Wear protection is a budget decision with a long tail. Unfilled PLA and PLA and PBAT blends are not abrasive, and nitrided screw elements with nitrided barrel liners perform well. Once calcium carbonate, talc, glass fiber or natural fiber enters the recipe at 20 percent or more, wear rates rise sharply and the calculation changes. Powder metallurgy screw elements in the high-shear kneading zones combined with bimetallic barrel liners in the same zones typically extend service life by a factor of three to five for an incremental spend of roughly 1 to 2 index points on the total project. The correct strategy is selective: premium materials in the two or three most aggressive barrel positions, standard nitrided components elsewhere.
| Recipe family | Abrasion level | Recommended screw material | Recommended barrel liner | Index point impact |
|---|---|---|---|---|
| Neat PLA, PLA/PBAT unfilled | Low | Nitrided alloy steel | Nitrided | Baseline |
| PLA + starch (thermoplastic starch) | Low–Medium | Nitrided, corrosion-resistant in vent zone | Nitrided | +0.3–0.6 |
| PLA + 20–40% calcium carbonate | Medium–High | Powder metallurgy in kneading zones | Bimetallic in mixing zones | +1.0–1.6 |
| PLA + talc + nucleating agent | Medium | Powder metallurgy selectively | Bimetallic selectively | +0.8–1.2 |
| PLA + natural fiber or wood flour | High | Powder metallurgy, full mixing section | Bimetallic, extended zones | +1.4–2.0 |
| PLA + glass fiber (engineering grade) | Very High | Powder metallurgy throughout | Bimetallic throughout | +1.8–2.6 |
Upstream Metering: Loss-in-Weight Feeders and Dosing Architecture
Feeding accuracy determines product consistency more directly than any other subsystem, and the dosing package absorbs 4 to 8 index points. In a starve-fed co-rotating twin-screw extruder, the feeders — not the screw speed — set the throughput and the recipe. If the feeder drifts, the recipe drifts, and no amount of downstream mixing will recover it.
Loss-in-weight feeders, also called gravimetric feeders, continuously weigh the hopper and adjust screw speed to hold a set mass flow rate. Well-implemented, they hold ±0.5 percent of setpoint over a 30-second averaging window. Volumetric feeders are cheaper and adequate for a single-component, free-flowing main resin stream, but for any minor component that affects certification — nucleating agent, chain extender, stabilizer, color masterbatch — gravimetric control is the only defensible choice.
The number of feeder positions is the main cost driver, and it should be decided from the recipe portfolio rather than from the current single product. A typical PLA and PBAT plant needs a main resin feeder, a second polymer feeder, one or two powder filler feeders, a masterbatch feeder and a liquid additive pump. Kerke supplies the complete family: volumetric metering systems, loss-in-weight feeders, twin-screw side feeders, crammer feeders for low bulk density powder, and liquid feeders with heated lines where the additive requires it.
Dosing Architecture for a Multi-Recipe PLA Plant
| Feed stream | Typical dose range | Feeder type | Injection point | Accuracy target |
|---|---|---|---|---|
| PLA base resin | 40–90% | Loss-in-weight, twin-screw metering | Main throat, barrel 1 | ±0.5% |
| PBAT or flexible co-polyester | 10–50% | Loss-in-weight, single-screw metering | Main throat, barrel 1 | ±0.5% |
| Mineral filler (calcium carbonate, talc) | 5–40% | Loss-in-weight + twin-screw side feeder | Side feed, after first melting zone | ±0.5% |
| Thermoplastic starch | 5–30% | Crammer feeder (low bulk density) | Side feed | ±0.7% |
| Nucleating agent / chain extender | 0.3–3% | Micro loss-in-weight feeder | Main throat or side feed | ±0.5% |
| Color or additive masterbatch | 0.5–5% | Micro loss-in-weight feeder | Main throat | ±0.5% |
| Liquid plasticizer or lubricant | 0.2–4% | Liquid feeder with metering pump | Melt zone injection port | ±1.0% |
Two practical budget notes. First, always specify one more feeder position than the current recipe needs. Adding a feeder later means new platform steel, new cable routing, a control system change and a production stoppage — collectively far more than the cost of the feeder itself. Second, the feeder platform structure, refill loaders and central material handling are frequently omitted from machine quotations and appear later as an unpleasant surprise. Ask explicitly whether the platform, stairs, guardrails and refill system are inside or outside the scope.
Midstream: Side Feeder, Devolatilization, Melt Pump and Screen Changer
The midstream package — side feeder, vacuum devolatilization system, melt pump and screen changer — absorbs 3 to 6 index points and is where PLA quality is either preserved or destroyed. These four components govern how much thermal and hydrolytic history the polymer accumulates between the feed throat and the die.
The twin-screw side feeder introduces powder filler downstream of the melting zone. Feeding mineral filler through the main throat with the resin would force the filler through the entire melting section, generating unnecessary shear heat and wear while risking bridging at the throat. Side feeding into an already-molten polymer stream also dramatically improves dispersion, because the filler agglomerates are broken down by kneading blocks acting on a low-viscosity carrier rather than on a bed of unmolten pellets.
Vacuum devolatilization removes residual moisture, lactide monomer and volatile decomposition products from the melt. For PLA, this stage is not cosmetic — it protects molecular weight. A single vacuum vent at 100 to 200 mbar absolute is the minimum; a two-stage arrangement with an atmospheric vent followed by a deep vacuum vent at 30 to 100 mbar absolute is strongly recommended where feed moisture control is imperfect or where recycled in-house scrap enters the recipe. A vapor condenser ahead of the vacuum pump keeps the pump oil clean and prevents oligomer deposits in the vacuum line, which is a maintenance item plants routinely underestimate.
The melt pump, sometimes called a gear pump, decouples extruder pressure from die pressure. It lets the extruder operate at low, stable head pressure — good for the polymer — while delivering high, extremely steady pressure to the die. For pellet size uniformity this matters: melt pressure pulsation translates directly into strand diameter variation and pellet weight variation. On a PLA line where a customer may specify pellet count per kilogram within a tight band, a melt pump moves from optional to necessary.
The screen changer protects the pelletizing die from contamination. For clean virgin PLA compounding a hydraulic plate screen changer is sufficient. Where post-industrial scrap or reprocessed material enters the feed, a continuous dual-piston screen changer is worth the extra index points, because it allows filter changes without stopping the line — and every unplanned stop on a PLA line means purging the barrel to avoid the thermal degradation that occurs when stagnant melt sits at temperature.
Pelletizing Routes Compared: Water-Ring, Underwater and Strand
Pelletizing route selection absorbs 4 to 8 index points and has a disproportionate effect on operating labor, pellet quality and moisture management. Three routes dominate PLA compounding, and each interacts differently with the polymer’s two weaknesses — thermal sensitivity and hydrolytic sensitivity.
Strand pelletizing extrudes strands through a die plate, cools them in a water trough, dries them with an air knife, and cuts them in a mechanical pelletizer. It is the simplest, most transparent and most tolerant route, and the operator can see immediately when a die hole plugs. Its weaknesses on PLA are the long water contact path and the manual strand start-up, which becomes awkward on high-output lines and generates start-up scrap.
Water-ring die-face hot cutting cuts the melt at the die face with rotating blades inside a housing where a ring of water flings the freshly cut pellets to the outlet. Melt contact with water is brief, so hydrolysis exposure is limited, and start-up is far easier than threading strands. This is the most widely selected route for PLA and PBAT compounding lines and is the default recommendation for Tier B and Tier C plants.
Underwater pelletizing cuts at the die face fully submerged in a temperature-controlled water chamber, then transports the pellets as a slurry to a centrifugal dryer. It produces the roundest, most uniform, lowest-dust pellet and is the preferred route for premium applications such as 3D printing filament feedstock, where pellet geometry consistency affects downstream melt stability. It carries the highest process water conditioning load and the highest control complexity. Kerke also offers air-cooled strand pelletizing, air-cooled die-face hot cutting and eccentric water mist hot cutting for recipes where water contact must be minimized further.
Pelletizing Route Comparison for PLA-Based Compounds
| Criterion | Strand pelletizing | Water-ring die-face cutting | Underwater pelletizing |
|---|---|---|---|
| Index points within pelletizing family | Low (baseline) | Medium (1.3–1.6× baseline) | High (1.9–2.6× baseline) |
| Melt-to-water contact time | Longest (trough path) | Short | Short but fully submerged |
| Hydrolysis risk on PLA | Medium | Low–Medium | Medium (needs tight water temperature control) |
| Pellet shape uniformity | Cylindrical, moderate spread | Lens-shaped, good | Spherical, excellent |
| Fines and dust generation | Medium–High | Low–Medium | Low |
| Start-up scrap | High (strand threading) | Low | Low |
| Suitability for high filler loading | Excellent | Good | Good |
| Suitability for soft, tacky PBAT-rich blends | Poor (strand sagging) | Good | Excellent |
| Operator skill requirement | Medium | Medium | High |
| Recommended tier | Tier A, filled rigid recipes | Tier B and Tier C, general purpose | Tier C and Tier D, premium grades |
One route decision cascades into the utility budget. Underwater pelletizing needs a closed-loop process water system with filtration, temperature control and level management, which adds pump, chiller and tank capacity to the utility family. Water-ring cutting needs a smaller circuit. Strand pelletizing needs the least. When comparing quotations, always confirm which package includes the water circuit, the dewatering centrifuge and the pellet dryer, because these are often quoted separately.
Drying and Dehumidification: The Line Item You Cannot Cut
Drying absorbs 4 to 7 index points and is the one budget family where cost reduction directly destroys product quality. PLA is a polyester and it hydrolyzes in the melt: water molecules attack ester linkages, chain length falls, molecular weight drops, and every mechanical property that matters follows it down. The damage is invisible at the extruder and only appears in the laboratory as a rising melt flow rate and a falling impact strength.
The target is straightforward. PLA resin must enter the extruder below 250 ppm moisture, and for high-molecular-weight grades or reactive extrusion the target tightens to below 100 ppm. Reaching those numbers requires a desiccant dehumidifying dryer delivering process air at minus 40 degrees Celsius dew point, drying temperature 60 to 80 degrees Celsius, and residence time of 4 to 6 hours in an insulated hopper sized for the line throughput. A simple hot-air dryer cannot reach these numbers because it recirculates ambient humidity.
Drying temperature has an upper bound that catches newcomers. PLA has a glass transition temperature in the 55 to 62 degrees Celsius range, so drying too hot causes pellets to soften, stick and bridge in the hopper. That is why the window is narrow: hot enough to drive out moisture, cool enough to avoid tackiness. Crystallized PLA grades tolerate the upper end of the range; amorphous grades demand caution and often a two-stage approach with a crystallizing stage first.
PLA Drying Parameter Window
| Parameter | Recommended value | Acceptable range | Failure mode if violated |
|---|---|---|---|
| Process air dew point | −40 °C | −35 to −45 °C | Residual moisture stays above target; chain scission |
| Drying temperature | 70 °C | 60–80 °C | Too high: hopper bridging; too low: incomplete drying |
| Drying time | 4–6 h | 4–8 h | Short: wet core pellets; long: thermal aging |
| Target inlet moisture | <250 ppm | <100 ppm for reactive grades | Molecular weight loss, MFR drift, brittle product |
| Airflow per unit throughput | 1.8–2.5 m³/h per kg/h | 1.5–3.0 | Uneven hopper temperature profile |
| Hopper insulation | Mandatory | — | Cold wall condensation re-wets pellets |
| Conveying line after dryer | Dry air or short closed path | — | Re-absorption within minutes in humid climates |
A detail that costs plants real money: dried PLA re-absorbs moisture quickly. Conveying dried resin through a long open-loop pneumatic line using humid ambient air can undo several hours of drying in a few minutes. Keep the path from dryer to feed throat short, closed and, in humid climates, purged with dry air. Also budget a second drying hopper if you run multiple recipes — switching recipes with a single hopper means either waiting out a full drying cycle or accepting a wet start.
Downstream Handling, Screening, Metal Detection and Packaging
The downstream package absorbs 3 to 9 index points and scales strongly with the automation level you choose. Everything after the pelletizer exists to guarantee that what leaves the plant is uniform, contamination-free, correctly labeled and traceable — requirements that biodegradable compound buyers enforce as strictly as any engineering plastic customer.
A vibrating classifier screen removes oversize agglomerates and undersize fines. On PLA lines this is more important than on commodity polyolefin lines, because slow crystallization means pellets can leave the dewatering stage still slightly soft and prone to doubles. A two-deck screen with an upper oversize deck and a lower fines deck is the standard configuration. Fines that pass the lower deck should be collected, not returned blindly to the feed, since repeated thermal history further degrades molecular weight.
Metal detection and magnetic separation sit immediately after screening. A magnetic drum or grate catches ferrous contamination; an inline metal detector with an automatic reject gate catches non-ferrous fragments. Blade fragments from the pelletizer, bolts and weld spatter from installation, and staples from raw material bags are all real sources. A single metal fragment reaching a customer’s injection molding machine screw or a film die will cost far more in claims and reputation than the detector.
Pneumatic conveying moves pellets from the screen to blending silos or directly to packing. Dilute-phase conveying is simpler and cheaper; dense-phase conveying is gentler and generates fewer angel hair streamers and fines, which matters for premium grades. Blending silos serve a quality function beyond storage: homogenizing several hours of production into one lot smooths out short-term process variation, so that the certificate of analysis reflects a genuinely uniform product.
Packing configuration should match the customer base. Small bag packing at 25 kilograms with a semi-automatic scale and sewing station suits diversified customers. Bulk bag packing at 500 to 1,250 kilograms suits volume customers with their own unloading stations. Most plants install both, and the choice weights the downstream index points accordingly. Add desiccant and moisture-barrier liners for any product shipped to a humid destination — a compound dried to 200 ppm and packed in a permeable bag will not arrive at 200 ppm.
Quality Laboratory and Compostability Testing Budget
The quality laboratory absorbs 2 to 5 index points and is proportionally most expensive for the smallest plant, because the instrument list is nearly fixed regardless of tonnage. A biodegradable compounder without an in-house laboratory cannot release product with confidence, cannot troubleshoot process drift, and cannot develop new recipes — so this family should never be cut, only phased.
Core Laboratory Instrument List
| Instrument | What it controls | Test frequency | Priority |
|---|---|---|---|
| Melt flow index tester (ISO 1133 method) | Molecular weight retention, batch consistency | Every lot, every 2–4 h | Phase 1 — essential |
| Moisture analyzer (coulometric or loss-on-drying) | Feed and product moisture in ppm | Every shift and every dryer batch | Phase 1 — essential |
| Universal tensile tester with extensometer | Tensile strength, elongation at break | Every recipe lot | Phase 1 — essential |
| Impact tester (Izod and Charpy) | Notched impact strength, toughening效果 verification | Every recipe lot | Phase 1 — essential |
| Differential scanning calorimeter | Glass transition, cold crystallization, melting behavior | Development and audit lots | Phase 1 for PLA |
| Ash content furnace | Actual filler loading verification | Every filled lot | Phase 1 for filled recipes |
| Colorimeter / spectrophotometer | Color consistency, yellowing index | Every colored lot | Phase 1 for masterbatch work |
| Laboratory twin-screw extruder | Recipe development without stopping the main line | Continuous R&D use | Phase 2 — high value |
| Heat deflection temperature apparatus | Thermal performance of nucleated grades | Development lots | Phase 2 |
| Capillary rheometer | Shear viscosity curve for processing windows | Development lots | Phase 3 — optional |
A Kerke laboratory twin-screw extruder deserves special mention as a budget item. It is designed for formula trials, tests and research work, and it lets a plant develop and screen recipes on small material quantities without interrupting production on the main line. For a merchant compounder serving multiple customers with different specifications, a laboratory extruder frequently pays for itself in avoided production line downtime within the first eighteen to twenty-four months of operation.
Compostability testing is a separate budget line and one that operates on a slow clock. Certification of a compound family against EN 13432 or ASTM D6400 involves disintegration and biodegradation testing that runs for months, with ecotoxicity and heavy metal screening alongside. Two planning consequences follow. First, start the testing process during construction, not after commissioning, because the laboratory clock runs in parallel with civil works at no schedule cost. Second, budget per recipe family rather than per product: a well-defined family with bounded composition ranges is more economical to certify than a scattered set of one-off formulations.
PLA-Specific Process Difficulties That Move the Budget
PLA is not a difficult polymer to compound, but it is an unforgiving one. Four material characteristics — a narrow thermal window, hydrolytic sensitivity, slow crystallization, and blend incompatibility with polyolefin-style thinking — drive specific equipment choices that a generic compounding line will not have. Understanding them explains why several index points in this budget go where they go.
Thermal Window: 170 to 210 Degrees Celsius
PLA melts around 150 to 175 degrees Celsius depending on stereochemistry and crystallinity, and thermal degradation accelerates above roughly 210 to 220 degrees Celsius. The practical processing window is therefore 170 to 210 degrees Celsius — narrow by the standards of polyolefin compounding. Two equipment consequences follow. First, barrel temperature control must be tight, which means well-tuned zone control and effective barrel cooling, not just heating. Second, shear heating must be managed by screw design rather than absorbed by lowering set temperatures, because there is very little headroom to absorb. This is exactly why the high-torque machine class matters: it delivers mixing energy at lower screw speed and lower melt temperature.
Hydrolysis and Residence Time Distribution
Every second that PLA spends hot is a second during which residual moisture can attack the chain. Residence time distribution — the spread of times different melt elements spend in the machine — therefore has a direct effect on molecular weight retention. A well-designed PLA screw has a narrow residence time distribution with no dead zones, no over-long kneading blocks, and no stagnant pockets in the die adapter. Typical mean residence time on a properly configured co-rotating twin-screw extruder running PLA is 25 to 60 seconds. Long-tail residence — material trapped behind a poorly matched transition or in an oversized adapter — produces the yellow specks and brittle particles that show up as customer complaints.
Slow Crystallization and Nucleation
PLA crystallizes slowly compared with polypropylene or polyethylene terephthalate. Left unnucleated, molded or extruded parts remain largely amorphous, giving low heat deflection temperature and a product that deforms in a hot warehouse or a dishwasher. Nucleating agents — talc at fine particle size, or specialty organic nucleators — raise crystallization rate and heat resistance, but they must be dispersed properly to work, which is a mixing task, not a dosing task. This is a case where dispersive mixing capability in the screw design directly determines whether a product specification can be met.
Blend Systems: PLA with PBAT, Starch and Mineral Filler
| Formulation family | Typical composition | Screw configuration emphasis | Target end products |
|---|---|---|---|
| Flexible film compound | PLA 20–40%, PBAT 45–70%, filler 5–20%, compatibilizer | Distributive mixing, gentle shear, low melt temperature | Compostable shopping bags, waste bags, mulch film |
| Rigid injection compound | PLA 70–90%, impact modifier 5–15%, talc 3–10%, nucleator | Dispersive mixing for nucleator, tight thermal control | Cutlery, cups, lids, rigid caps |
| Starch-based blend | PLA 30–50%, thermoplastic starch 20–40%, PBAT balance | Strong devolatilization, crammer side feeding | Low-cost bags, agricultural film, food trays |
| High-filler rigid compound | PLA 50–70%, calcium carbonate 25–40%, coupling agent | Side feeding, wear-protected mixing zone | Thermoformed trays, rigid sheet, disposable tableware |
| Sheet extrusion compound | PLA 80–95%, chain extender 0.3–1%, processing aid | Melt strength building, deep vacuum devolatilization | Packaging sheet, blister and clamshell packaging |
| Filament-grade compound | PLA 85–98%, toughener, color masterbatch | Fine melt filtration, uniform pellet geometry | 3D printing filament feedstock |
PLA and PBAT are only partially compatible. Without a compatibilizer, the blend forms a coarse dispersed phase with poor interfacial adhesion, and the film tears. With a reactive compatibilizer and adequate distributive mixing, the dispersed phase refines to a scale where the blend behaves as a tough, drawable material. This is a compounding problem, not a formulation problem alone — the same recipe run on a poorly configured screw will not produce the same film. It is also the strongest argument for buying compounding capability rather than outsourcing: a converter who controls the screw configuration controls the property envelope of the finished product.
PLA resin selection follows the application. Injection-grade PLA with a melt flow rate of roughly 6 to 20 g/10 min suits rigid molded articles; extrusion and sheet grades sit lower, in the 2 to 8 g/10 min band, for melt strength; film grades and filament feedstock sit in between. Specify by generic grade family and property window rather than by a fixed supply source, so that the plant can qualify alternative feedstock without re-engineering the process.
Kerke Machine Blocks: KTE Series Line Configurations
Kerke builds the KTE series of parallel co-rotating twin-screw extruders from KTE-16B through KTE-135D, covering outputs from around 30 kg/h on the smallest laboratory-scale machines up to high-capacity industrial lines. For PLA compounding, three size classes cover the four capacity tiers described earlier. Each machine is configurable in aspect ratio, barrel structure, screw arrangement, exhaust configuration, feeding architecture and electrical control, so the specification below is a representative PLA-duty build rather than a fixed catalog entry.
Kerke KTE-52D — Tier A and Tier B PLA Compounding Line
The KTE-52D is the workhorse for a first commercial biodegradable compounding line. It has enough torque and L/D to handle a PLA and PBAT blend with side-fed mineral filler and a two-stage vacuum train, while remaining compact enough for a modest building envelope and a mid-size transformer. For a Tier A or Tier B project, this is the machine that produces the best balance of capability and capital intensity per annual ton.
| Specification — KTE-52D (PLA duty build) | Value |
|---|---|
| Screw diameter | 51–52 mm, co-rotating parallel twin-screw |
| L/D ratio | 40:1 standard, extendable to 48:1 |
| Main motor power | 55–90 kW, AC variable frequency drive |
| Maximum screw speed | 600–800 rpm |
| Specific torque grade (T/A³) | 8–11 N·m/cm³ (high-torque class) |
| Output range on PLA compounds | 150–400 kg/h depending on recipe and filler level |
| Barrel construction | Segmented, bolted; closed / side-feed / vent types; nitrided or bimetallic liner |
| Barrel heating and cooling | Electric heating with forced-air or water cooling, independent zone control |
| Devolatilization | Atmospheric vent plus vacuum vent, 30–100 mbar absolute with condenser |
| Feeding package | 2–4 loss-in-weight feeders, twin-screw side feeder, optional liquid feeder |
| Pelletizing | Water-ring die-face hot cutting or water-cooled strand pelletizing |
| Control system | PLC with touch-screen HMI, recipe management, data acquisition |
| Recommended capacity tier | Tier A (500 t/y) and Tier B (1,000 t/y) |
Kerke KTE-75D — Tier C Merchant Compounding Line
The KTE-75D steps up to a genuine merchant compounding scale. The larger free volume accommodates higher filler loadings and a longer, more differentiated screw profile, so a single machine can serve several recipe families — flexible film compound in one campaign, rigid nucleated compound in the next — without a physical rebuild between them. For plants intending to sell compound to third-party converters rather than only feeding internal lines, this is usually the correct starting size.
| Specification — KTE-75D (PLA duty build) | Value |
|---|---|
| Screw diameter | 71–73 mm, co-rotating parallel twin-screw |
| L/D ratio | 44:1 standard, extendable to 48:1 |
| Main motor power | 160–250 kW, AC variable frequency drive |
| Maximum screw speed | 600–800 rpm |
| Specific torque grade (T/A³) | 8–12 N·m/cm³ (high-torque class) |
| Output range on PLA compounds | 500–1,200 kg/h depending on recipe and filler level |
| Barrel construction | Segmented, bolted; bimetallic liner in mixing zones for filled recipes |
| Devolatilization | Two-stage vacuum, 30–100 mbar absolute, condenser and separator |
| Melt pump | Recommended, for pellet uniformity and stable die pressure |
| Screen changer | Hydraulic plate or continuous dual-piston type |
| Feeding package | 4–6 loss-in-weight feeders, twin-screw side feeder, crammer feeder option, liquid feeder |
| Pelletizing | Water-ring die-face hot cutting or underwater granulation system |
| Recommended capacity tier | Tier C (2,000 t/y), or one of two lines at Tier D |
Kerke KTE-95D — Tier D Industrial Line
For a 5,000 ton per year installation the choice is between two KTE-75D lines and a single larger machine such as the KTE-95D. Two lines give redundancy and recipe flexibility; one large line gives lower operating labor per ton and lower energy intensity. The KTE-95D configuration below suits a plant with a narrow, stable product range and high volume per recipe.
| Specification — KTE-95D (PLA duty build) | Value |
|---|---|
| Screw diameter | 92–95 mm, co-rotating parallel twin-screw |
| L/D ratio | 44:1 to 48:1 |
| Main motor power | 315–500 kW, AC variable frequency drive |
| Maximum screw speed | 500–700 rpm |
| Specific torque grade (T/A³) | 8–12 N·m/cm³ |
| Output range on PLA compounds | 1,000–2,500 kg/h depending on recipe |
| Devolatilization | Two- or three-stage vacuum with independent pump sets |
| Pelletizing | Underwater granulation system with centrifugal dryer |
| Typical energy intensity | 0.22–0.30 kWh per kg of compound |
| Recommended capacity tier | Tier D (5,000 t/y) single-line configuration |
Beyond the standard KTE series, Kerke offers three configurations worth knowing when a recipe resists a conventional single-stage approach. The KTE-T triple-screw extruder handles special materials where the three-screw geometry improves distributive mixing and surface renewal. The KTE-SE double-stage extrusion system — a mother-and-baby arrangement in which a second extruder receives melt from the first — suits materials that cannot be processed on a one-stage machine, typically because devolatilization demand or shear sensitivity is extreme. And the SE series single-screw extruder, with outputs from 30 kg/h to 800 kg/h, serves reprocessing duty on in-house scrap. A plant that expects to reclaim its own edge trim and off-spec production should budget one from the start rather than sending scrap out.
Application Industries and End Products
Kerke’s compounding installations serve masterbatch production — color masterbatch, filler masterbatch, additives masterbatch, black masterbatch and textile masterbatch — and plastic compounding across engineering plastics, biodegradable plastics, cable compounding, PVC compounding, thermoplastic elastomers and wood-plastic composites, as well as R-PET flake recycling and specialized food processing extrusion. For a PLA plant specifically, the compound leaving your pelletizer will go into compostable shopping bags and waste bags, agricultural mulch film, disposable cutlery, cups and lids, thermoformed food trays and clamshell packaging, rigid packaging sheet, coffee capsules, and 3D printing filament feedstock. Each of those end products imposes a slightly different property target, which is why recipe flexibility — feeder count, L/D headroom, screw modularity — is worth more than a marginal increase in nameplate throughput. For downstream conversion equipment, Wanplas supplies matched lines that integrate with Kerke compounding systems, so a compounder that later moves into film or sheet conversion can extend within one supply relationship.
Capacity and Formulation to Model Selection
Model selection follows from three inputs: target annual capacity, formulation family, and filler loading. The table below maps common PLA project profiles to the recommended Kerke configuration. Treat it as a starting point for discussion rather than a final specification, because pellet geometry requirements, in-house scrap policy and future recipe plans all shift the answer.
| Target annual capacity | Formulation system | Filler loading | Recommended Kerke model | Key configuration notes |
|---|---|---|---|---|
| Trials and R&D only | Any PLA system | 0–40% | Laboratory twin screw extruder | Small batch capability; scale-up geometry matched to production models |
| 300–600 t/y | PLA/PBAT film compound | 5–20% | KTE-52D, L/D 40 | Water-ring cutting; 3 loss-in-weight feeders; single vacuum stage |
| 800–1,200 t/y | PLA/PBAT plus mineral filler | 20–35% | KTE-52D, L/D 44–48 | Twin-screw side feeder; bimetallic liner in mixing zones; two vacuum stages |
| 800–1,200 t/y | Rigid nucleated PLA for cutlery and cups | 3–10% talc | KTE-52D or KTE-65D | Dispersive kneading emphasis; tight barrel temperature control; melt pump |
| 1,500–2,500 t/y | Multi-recipe merchant compounding | 0–40% | KTE-75D, L/D 44–48 | 6 feeder positions; continuous screen changer; melt pump; blending silos |
| 1,500–2,500 t/y | Filament-grade and premium pellet geometry | 0–10% | KTE-75D with underwater granulation | Fine melt filtration; process water temperature control; centrifugal dryer |
| 4,000–6,000 t/y | Narrow product range, high volume | 10–30% | KTE-95D single line | Lowest energy intensity and labor per ton; bulk silo feeding |
| 4,000–6,000 t/y | Broad recipe portfolio, frequent changeover | 0–40% | Two KTE-75D lines | Redundancy; parallel campaigns; simpler recipe scheduling |
| Any tier, shear-sensitive system | Heavy devolatilization or reactive extrusion | Variable | KTE-SE double-stage system | Second-stage extruder decouples mixing from pressure build and pelletizing |
| Any tier, scrap reclaim | In-house edge trim and off-spec pellets | — | SE series single screw extruder | 30–800 kg/h; keeps reclaim off the main compounding line |
Plant Layout, Manpower and Energy Intensity
Layout, manpower and energy intensity are not separate budget lines, but they determine how many index points you actually earn back. A poor layout raises labor, raises scrap, and raises energy; a good single-direction flow pays for itself continuously. The canonical compound flow is strictly linear: raw material warehouse to drying, drying to metering, metering to extrusion, extrusion to pelletizing, pelletizing to screening, screening to metal detection, detection to conveying, conveying to blending and packing, with finished goods separate from incoming resin. Cross-flows — a forklift carrying bags across the pellet path, a dryer return line crossing the lab — invite contamination and accidents.
Personnel Complement by Shift Pattern
Manning is driven by the number of operating lines and the degree of automation in packing and material handling. The figures below are operator-level roles; they exclude management, quality and maintenance supervision, which add roughly 25 to 40 percent on top depending on plant maturity.
| Role | Tier A (1 shift) | Tier B (2 shifts) | Tier C (3 shifts) | Tier D (3 shifts, 2 lines) |
|---|---|---|---|---|
| Line operators | 2 | 4 | 6 | 10–12 |
| Material handling and packing | 1–2 | 3 | 4–5 | 8–10 |
| Quality and laboratory | 1 | 1–2 | 2–3 | 3–4 |
| Maintenance technician | 0.5 (shared) | 1 | 1–2 | 2–3 |
| Supervision and logistics | 0.5 (shared) | 1 | 1–2 | 2–3 |
| Total direct headcount | 5–6 | 10–11 | 14–17 | 25–32 |
Specific Energy Consumption by Machine Class
Energy intensity is the operating cost that most directly competes with the capital budget. A machine with a marginally lower purchase price but a higher specific energy consumption will, over a ten-year life, consume far more in electricity than was saved up front. The index point model rewards choosing the efficient machine.
| Machine class | Typical kWh per kg compound | Relative grade | Main levers |
|---|---|---|---|
| KTE-52D, unfilled PLA/PBAT | 0.28–0.38 | Medium | Drive efficiency, motor sizing to load |
| KTE-75D, 30% filler | 0.25–0.32 | Medium | Torque utilization, barrel cooling duty |
| KTE-95D, narrow range | 0.22–0.30 | Low | Scale effect, near-continuous running |
| Single-screw reclaim line (SE series) | 0.15–0.22 | Low | Lower shear, lower throughput duty |
| Drying and dehumidification load | 0.04–0.08 | Low–Medium | Desiccant regeneration cycle control |
| Pellet water circuit and dewatering | 0.03–0.06 | Low | Pump speed matching to flow |
A useful planning figure: at typical industrial electricity rates, a machine class that is 0.05 kWh/kg better than an alternative, running 6,000 tonnes per year, avoids enough energy to pay a noticeable share of one index point over the plant life. Energy efficiency is therefore not only an operating-cost concern but a budget-allocation concern, and it belongs in the machine specification scorecard alongside torque and L/D.
Construction Schedule: Month-by-Month Gantt
The project schedule is the bridge between the index point model and reality. A compounding plant from greenfield to stable production typically runs twelve to twenty months, and the long-pole items — civil works and utility grid connection — set the critical path. Treating commissioning as a separate phase rather than an afterthought is what separates a plant that reaches nameplate in month sixteen from one that struggles to month twenty-four.
Reference Construction Schedule (Tier B, 1,000 t/y single line)
| Phase | Month 1–2 | Month 3–5 | Month 6–8 | Month 9–11 | Month 12–14 |
|---|---|---|---|---|---|
| Design and permitting | Active | Close-out | — | — | — |
| Civil works and structure | Start | Active | Complete | — | — |
| Utility grid and transformer | Apply | Install | Energize | — | — |
| Equipment manufacture and FAT | Order | Build | Factory test | Ship | — |
| Installation and erection | — | — | Begin | Active | Complete |
| Single-machine commissioning | — | — | — | Start | Active |
| Integrated trial run | — | — | — | — | Active |
| Start of production | — | — | — | — | Month 13 |
| Ramp to nameplate and certification | — | — | — | — | Month 14–18 |
Two scheduling lessons from the table. First, the equipment factory acceptance test should happen while civil works are still finishing, so that any mechanical correction is done in the supplier’s shop rather than on your floor. A factory acceptance test on a PLA line typically runs continuous operation for several shifts on a representative recipe, confirming output, energy intensity, melt temperature stability and pellet quality before shipment. Second, start compostability testing of your lead recipe family in month three or four. Because that clock is months long, running it in parallel with construction compresses time-to-certification by a full season.
Budget Risk Ranking and Overrun Control
Every budget has overrun candidates. Ranking them by probability and magnitude lets you hold a contingency reserve in the right places instead of spreading it thin. The table uses relative grades — Low, Medium, High, Very High — and expresses the potential swing as index point ranges on the 100 point total.
| Risk item | Probability | Potential swing | Mitigation |
|---|---|---|---|
| Utility grid connection delay | High | +0 to +3 index points (schedule, not capex) | Apply in month 1; build temporary generator contingency |
| Soil condition requiring deeper foundations | Medium | +1 to +3 index points | Geotechnical survey before final civil design |
| Scope creep in feeder and platform count | High | +1 to +4 index points | Fix feeder count from full recipe portfolio, not current product |
| Under-specified pellet water circuit | Medium | +1 to +2 index points | Scope chiller, filtration and dryer in main quotation |
| Emission treatment upgrade after inspection | Medium | +1 to +3 index points | Pre-clear permit conditions with local authority early |
| Wear package under-bought for filled recipes | Low–Medium | +1 to +2 index points | Specify selective bimetallic for the worst recipe |
| Trial material loss above plan | Medium | +1 to +2 index points | Reclaim on single-screw line; minimize recipe count at start |
| Certification retest for non-compliant lot | Low–Medium | +0.5 to +1.5 index points | Lock formulation range; in-house DSC and MFR control first |
| Working capital underestimation | High | +1 to +3 index points | Model resin inventory at 30–45 days; include payment terms gap |
The single most damaging overrun is not in any equipment line — it is a schedule slip caused by the utility grid connection running long. A delayed energization pushes commissioning, which pushes the integrated trial run, which pushes first revenue, and the cost of that delay dwarfs the few index points saved by skimping elsewhere. Hold contingency in schedule first, then in the feeder/platform scope and the working capital, which together account for the majority of real overruns.
Phased Construction: Line One Now, Line Two Later
Phasing is the capital-allocation decision that protects option value. A phased build puts one line into production as early as possible, generates cash and learning, and adds a second line when the market justifies it. The alternative — building the full two-line plant on day one — commits the entire 100 index points before any market signal confirms demand. For a first-time PLA plant, phasing is almost always the lower-risk path.
Phasing Decision Guide
| Decision point | Build full plant now | Phase one line, reserve interface |
|---|---|---|
| Capital committed on day one | 100 index points | 60–70 index points |
| Time to first revenue | Same as single line | Same; earlier cash generation |
| Building envelope | Full footprint built | Full footprint shell built; second bay fitted later |
| Transformer and cooling | Sized for two lines | Sized for two lines but energized for one |
| Material handling | Full system | Single line now; second conveyor interface reserved |
| Risk if demand disappoints | Very High | Medium |
| Penalty for later expansion | None | +3 to +6 index points vs full build (duplicate interface work) |
The phased approach costs a few extra index points when you eventually expand, but it avoids committing the full 100 index points against an unproven demand forecast. The key is to reserve the interface on day one: build the building shell large enough, size the transformer and cooling header for two lines, and lay the foundations and cable trays for the future line. Those reservations are cheap; retrofitting them later is expensive. A well-designed Kerke line layout leaves clear space and utility stubs for a second machine without interrupting the first line’s operation.
Service, Commissioning and Support
The final budget families — first-fill spare parts, installation and commissioning, training, trial material and working capital — together represent a meaningful share of the 100 index points, and they are the difference between a machine that sits in a building and a plant that makes saleable compound. Kerke, as a Wanplas factory, treats these as part of the deliverable, not as optional extras.
Before shipment, each line undergoes factory testing and a factory acceptance test that confirms mechanical run-out, drive performance, control logic, and on representative material where possible, output and pellet quality. This testing catches the majority of commissioning problems in the supplier’s workshop rather than on the customer’s floor. On arrival, Kerke engineers supervise installation and commissioning, including alignment of the extruder and pelletizer, instrument calibration, feeder mapping, and the first production runs of the lead recipes.
Spare parts policy is stated plainly and applies to the whole Wanplas group: USD 500 free parts/year. This covers a defined set of wear and consumable items within that annual value, so that routine maintenance does not immediately become an unbudgeted cost. Combined with a warranty that includes free replacement of damaged parts within its term, the policy reduces the risk that a small component failure stalls a line. A replacement policy where quality failure triggers a refund plus compensation sits alongside this to protect the buyer’s downside on the machine itself.
Training covers three audiences: machine operators on start-up, shutdown and purge discipline specific to hydrolytically sensitive PLA; maintenance technicians on screw pull, barrel inspection and wear-part rotation; and quality staff on the in-house test methods — melt flow rate, moisture, impact, ash, and differential scanning calorimetry interpretation — that keep a lot release defensible. Remote monitoring and remote support let Kerke engineers diagnose many faults without a site visit, reducing unplanned downtime after handover.
Kerke operates an open-factory policy and welcomes customer visits for inspection and audit. For a biodegradable compounder, an in-person factory visit before order placement is the fastest way to verify the machine build, the control system, the feeding architecture and the supplier’s real experience with PLA recipes. Bring a sample of your intended resin and additive system; a meaningful supplier can often run a small trial on a laboratory extruder during the visit and show you pellet quality on the spot.
Frequently Asked Questions
How many index points of a PLA compounding plant budget go to the extruder itself?
In the normalized 100 index point model, the co-rotating twin-screw extruder and its integrated feeding, devolatilization and pelletizing package typically absorbs 28 to 35 index points. Civil works and utilities together usually take a similar share, which surprises first-time investors who assume the machine dominates the budget. The main machine alone, without feeders and pelletizing, sits at 16 to 24 index points depending on tier.
Why does PLA need dehumidifying drying before compounding?
PLA hydrolyzes rapidly in the melt phase when residual moisture is present. Feeding resin above roughly 250 ppm moisture causes measurable molecular weight loss, visible as a rising melt flow rate and falling impact strength. A dehumidifying dryer holding minus 40 degrees Celsius dew point at 60 to 80 degrees Celsius for 4 to 6 hours is a mandatory line item, not an optional accessory, and it occupies 4 to 7 index points.
Which pelletizing route suits a PLA and PBAT compounding line best?
Water-ring die-face hot cutting is the most common choice for PLA and PBAT blends because it combines short melt residence time with limited water contact. Underwater pelletizing gives the roundest pellets and best dust control but adds process water conditioning load, while strand pelletizing is the lowest-complexity route and remains attractive for stiff, high-filler formulations. The route choice shifts the pelletizing family between a baseline cost and up to 2.6 times that baseline.
What L/D ratio and torque class should a PLA compounding extruder have?
For a recipe that includes side-fed filler, liquid injection, devolatilization and pressure build, the L/D ratio should be 40 to 48. The specific torque class should be the high-torque range of 8 to 12 N·m/cm³, because PLA must be processed cool and you cannot simply raise melt temperature to reduce drive load. Choosing a 32 to 36 L/D standard-torque machine to save index points permanently restricts the recipe portfolio.
How long does it take to build and commission a PLA compounding plant?
A single-line plant of around 1,000 tonnes per year typically moves from design to first production in twelve to fourteen months, with ramp to nameplate over the following four to six months. The critical path is civil works and the utility grid connection, not the machine delivery, which is why the transformer application should be filed in the first month. Compostability certification against EN 13432 or ASTM D6400 should start during construction because its clock runs for months.
Should a first PLA plant build one line or two?
For a first-time plant, phasing is usually lower risk: build one line and reserve the interface for a second. This commits 60 to 70 index points on day one instead of the full 100, and generates revenue and operational learning earlier. The deliberate reservations — building shell, transformer capacity, cooling header, foundations and cable trays for the future line — cost little now and avoid expensive retrofitting later.
What energy intensity should I expect from a PLA compounding line?
Specific energy consumption on PLA compounds runs roughly 0.22 to 0.38 kWh per kilogram depending on machine size and filler loading, with larger machines at the lower end of the range. Drying adds 0.04 to 0.08 kWh per kilogram and the pellet water circuit adds 0.03 to 0.06. A machine class that is 0.05 kWh per kilogram better can offset a noticeable share of capital over the plant life, so energy efficiency belongs in the selection scorecard.
Does Kerke provide the downstream conversion lines as well as the compounding extruder?
Kerke supplies the compounding extruder, feeding, devolatilization and pelletizing island, plus a laboratory extruder and single-screw reclaim line from the same portfolio. For downstream conversion equipment, Wanplas supplies matched lines that integrate with Kerke compounding systems, so a compounder moving into film or sheet conversion can extend within one supply relationship rather than sourcing separate machinery.
Conclusion
Planning the total construction and equipment budget for a PLA twin screw compounding factory is a capital-allocation exercise before it is a machinery procurement exercise. Normalize the investment to 100 index points, allocate them deliberately across fourteen budget families, and the decisions that actually move the total — capacity tier, civil and utility scope, feeder and platform count, drying, pelletizing route, and working capital — become visible and controllable. The extruder is the technical heart of the plant but, by itself, only 16 to 24 index points; the building, the utilities, the upstream metering, the drying and the downstream handling collectively decide whether the project reaches nameplate on schedule.
Kerke, a Wanplas factory with 12 or more years in twin-screw compounding and more than 2,000 machines running in 70 or more countries, builds the KTE series co-rotating parallel twin-screw extruders that anchor this kind of plant — from the KTE-52D for a first commercial line through the KTE-75D for a merchant compounder to the KTE-95D for industrial volume, with laboratory, double-stage and reclaim options for the full recipe lifecycle. For a PLA and PBAT portfolio specifically, the high-torque, long L/D configuration with selective wear protection and a two-stage vacuum train is the configuration that preserves molecular weight while dispersing filler and nucleating agent properly.
If you are evaluating a PLA compounding investment, bring your target annual capacity, your intended formulation families and your filler loading to the discussion. Kerke can map them to a concrete KTE configuration, lay out the single-direction plant flow, size the civil and utility scope, and reserve the phasing interface for a second line when your market is ready. An open-factory visit is the most efficient way to verify the build, and a small trial of your own resin on a laboratory extruder will show pellet quality before you commit the full budget. We invite you to send your specification and arrange a factory audit — the compounding line that fits your product, your site and your growth plan is a configuration conversation, not a catalog pick.
