From 20 g to Under 8 g Why lightweighting and high-cavitation have become the whole ballgame in PET preform molding

Two levers, and not much else

Strip away the marketing and a preform plant has only two levers that materially change its cost per thousand pieces:

How much material is in each preform, and how many preforms leave the mold per hour.

Everything else — automation, digitalisation, service, spare parts — either supports those two numbers or is noise. In 2026 both levers have been pushed hard, and the interesting question is no longer whether they can be pushed further. It is what breaks first when you do.

1. Where lightweighting actually stands

The headline number came out of the PETnology Europe 2025 industry panel: preform weights have fallen from 20 g to under 8 g. The same discussion drew a sharper conclusion — the next round of reduction "requires a systems approach, not just new machines."

That distinction matters. The first 12 grams came off by changing the preform. The next gram comes off by coordinating the whole line: melt quality, mold cooling, take-out timing, and bottle performance validation.

Today's commercial preform weights, as listed by preform suppliers and mold makers, look roughly like this:

Read those numbers as market indications, not standards. Preform weight is a design decision driven by the bottle's top-load requirement, the fill temperature, the carbonation level, and how much the brand is willing to spend on validation.

2. Why the next gram is the hardest

Lightweighting is not a linear process. Once a preform gets thin, several constraints arrive at once:

Top load and stacking strength. A thinner wall means the filled, stacked bottle deforms. This is usually the binding constraint for still water and for anything palletised and warehoused.

Base design and stress cracking. The base is where wall distribution is hardest to control and where environmental stress cracking shows up. Lightweighting the base without changing base geometry is how field failures get designed in.

Carbonation. Carbonated soft drink bottles cannot follow still-water bottles down the weight curve — internal pressure dictates a heavier, petaloid-base design.

Mould and material limits. Mould makers are now quoting wall thickness as low as 1.6 mm, with micro-pitching for smaller neck sizes and neck ranges from roughly 12 mm to 140 mm. Each step below that asks more of both the mould and the machine's injection repeatability.

Neck finishes are effectively fixed. The neck has to match cap standards and capping equipment, so it is the one part of the preform that lightweighting cannot touch. All the reduction has to come out of the body and base.

The practical consequence: lightweighting has moved from a design exercise to a process-capability exercise. Two plants can run the same preform drawing and get different results, because the limiting factor is now how tightly each one can hold cavity-to-cavity wall thickness.

3. High cavitation: what the numbers look like

Cavitation has climbed steadily — 48 was a large tool a decade ago; 96, 128 and 144 are now standard offerings in the high-output segment. Mold makers advertise moulds up to 144 cavities with delivery times measured in weeks for volume neck sizes.

Here is what published line data looks like, next to the arithmetic:

Cycle times as published by mould and machine suppliers for high-cavitation PET preform lines (typically 10–12.8 s). Output columns are calculated, not quoted — and they are theoretical. Real lines run an OEE below 100%, so treat these as ceilings.

The five-row takeaway from that table: going from 96 to 144 cavities adds about 50% more output — it does not double it, because the heavier cavity count usually forces a slightly longer cycle. Cavitation is not free output. It is a trade.

4. What actually breaks when you scale cavities

Ask a supplier what happens between 48 and 144 cavities and you will hear about clamping force. In practice, the failure modes arrive in this order:

1. Shot weight versus machine capacity. A 144-cavity tool at 11 g needs a shot of about 1.58 kg per cycle, injected in a very short time. The injection unit, not the clamp, is usually the first genuine limit.

2. Hot runner balance. With 96 to 144 drops, keeping every cavity at the same filling pressure and the same thermal history is the hardest engineering problem in the tool. Unbalanced flow shows up as cavity-to-cavity wall thickness variation — which is exactly what lightweighting removed your margin for.

3. Mould rigidity and thermal uniformity. A larger tool deflects more under injection pressure and is harder to cool evenly. Non-uniform cooling produces non-uniform shrinkage.

4. Cooling time. In-mold cooling dominates the cycle. A 144-cavity tool has more heat to remove per cycle in the same window, so cooling moves from a design detail to the cycle-time gate.

5. Take-out time. More preforms per cycle means a longer, more complex take-out. This is where high-cavitation lines quietly lose the output they gained — unless the robot is engineered for it.

5. The bottleneck nobody prices in: cooling and take-out

This is the part of high-cavitation investment most frequently undervalued.

If the preform's cooling is completed inside the mold, the cycle time is set by how long the material needs to become rigid enough to eject without distortion. Post-cooling take-out systems break that dependency: the preform is removed earlier and finishes cooling outside the mould, in a robot-carried cooling station.

Suppliers now routinely pair high-cavity tools with:

multi-station take-out robots

• that grip, carry and cool the preforms in a secondary circuit

circulating-water cooling

• in the take-out path

centralised control

• coordinating robot, mold and machine

Published line examples for the 72–144 cavity segment describe three-station high-speed take-out robots with circulating-water cooling, and double-stage post-cooling robots designed specifically to take cooling load out of the mould.

The commercial point: a take-out robot is not an accessory. On a high-cavity line it is a cycle-time component. Buy it as part of the engineered system, with the cycle-time guarantee attached to the whole line — not to the machine alone.

6. The cost arithmetic that settles most arguments

Here is why lightweighting keeps winning even in years when energy prices dominate the headlines. Run the numbers on one 96-cavity line at 15 g preforms:

Assumptions: 432 kg/hour · 7,000 operating hours/year · resin at RMB 8,800/tonne (Chinese bottle-chip price, September 2026) · electricity at RMB 0.70/kWh

One percent of preform weight is worth roughly fifty times one percent of energy consumption.

This is an illustrative calculation from the stated assumptions, not a quotation. But the order of magnitude is the point — material cost dominates every other operating cost in a preform plant.

That single ratio explains the industry's behaviour: lightweighting programmes get funded even when they are technically difficult, and energy efficiency gets funded when it comes bundled with something else.

7. Energy is still worth measuring — properly

Energy matters, just not for the reason most brochures suggest. It matters because it is a proxy for how hard the machine is working, and because the spread between good and poor systems is wide.

The only energy figure worth comparing between suppliers is energy per kilogram of PET produced — not installed power, not "up to X% savings." A documented example: SIPA's XForm 500 was specified at under 220 W/kg of PET, i.e. about 0.22 kWh per kg, and was reported to have run 11 million cycles without maintenance.

Converted to the line above, the gap between a 0.22 kWh/kg system and a 0.30 kWh/kg system is about 242,000 kWh per year — roughly USD 24,000 at USD 0.10/kWh, on a single line. Meaningful, and worth negotiating for. But noticeably smaller than the material saving available from a 1% weight reduction on the same line.

Husky's 2025 platform launch took the same idea in a different direction: a fully integrated melt delivery system, robot and auxiliaries with an 8-cavity mould and side-entry robot — a configuration aimed at flexibility and changeover rather than maximum cavities. Both strategies are legitimate; they serve different business models.

8. It is a system, not a machine

The PETnology panel's conclusion applies directly to capital planning. Once you are chasing the last gram and the last second, the machine is one component in a chain:

Evaluation matrix — what to actually ask for

9. Where this goes next

Non-beverage is the growth pocket. Household chemicals, edible oil, dairy, personal care and pharma preforms carry wider weight tolerance but tighter regulatory and barrier requirements — a different competitive game from water bottles.

Barrier and multilayer preforms remain the answer for oxygen-sensitive products, and they raise the bar on co-injection control and hot runner design.

Market size, for context: depending on how the scope is drawn, the PET preform market is put at around USD 18.6 bn in 2025 growing to roughly USD 31.4 bn by 2034 (Dataintelo), while the narrower PET preform machines segment is estimated at USD 8.38 bn in 2022 rising to USD 14.73 bn by 2032 (Acumen Research & Consulting). Definitions vary between research houses — treat any single figure as indicative of direction, not of precision.

The takeaway

Lightweighting and high-cavitation are no longer separate projects. They are the same project: extracting material and time out of a process whose remaining margin lives in cooling, melt consistency and cavity-to-cavity repeatability.

The plants that win the next round will not be the ones with the most cavities on the nameplate. They will be the ones that can tell you, for every gram they remove, exactly what happened to wall thickness spread, cycle time and top-load performance.

Three questions worth asking any supplier before you commit capital:

1. What is the guaranteed cycle time for this preform weight in this cavity count, with take-out and post-cooling included?

2. What is the measured cavity-to-cavity wall thickness spread, and how was it measured?

3. What is the energy consumption in kWh per kg of PET, and where is it measured?