Mould Preheat Is Where Most Job Changes Lose Time on the Hot End

It's 4am on a Tuesday and the mould setter on section six is racing a preheat oven that never got the dwell time it needed. The gaffer's watching gob distribution drift on the monitor. The operator's already seeing checks on the first ware and nobody in the room wants to call it that yet.

Everyone blames the machine. It's rarely the machine. Mould preheat and seasoning is stage three of the nine-stage Job Change Lifecycle, and it's the stage where plants lose the most time without ever writing it down as lost time.

Cross-crew variance starts at the oven, not the IS machine

On paper, every crew runs the same preheat curve. In practice, night shift is racing the changeover clock and the day crew has more hands free around the oven. That gap is exactly where cross-shift variance on identical SKUs runs 30-60% in plants that haven't standardised the changeover, and it's rarely the mould's fault. It's whoever pulled it, and when, and off which reading.

I've walked into plants where three crews had three different unwritten rules for the same job. One waited the full dwell. One split the difference under schedule pressure. One just went by feel. Same mould, same SKU, three different first-run reject counts.

The temperature band nobody enforces consistently

Mould preheat target before installation on the IS machine sits at 450-565°C, roughly 850-1050°F. Moulds installed outside that band are the single most common root cause of first-run checks, thermal-shock cracks on the parison contact surface that scrap a gob before the line has even settled.

I once stood at a line and watched a mould go in around 20°C under spec because the oven display was reading ambient, not mould body temperature (nobody had recalibrated the thermocouple in over a year, and yes, the fitter swore it was fine, we checked it anyway). Standard work for gas-ring or oven rotation calls for a minimum dwell of 90-120 minutes to homogenise mould body temperature before installation. Moulds pulled early because the line is "waiting" are a common, informally tolerated deviation. It looks like a scheduling call. It's actually a process defect, and it's one most audits never flag because nobody logs the early pull anywhere. That's a gap the Job Change Lifecycle framework closes by making dwell time a locked, visible checkpoint instead of a judgement call.

Seasoning cycles are the corner everyone cuts first

Standard practice calls for three to five seasoning cycles, graphite or water-based release agent, run at reduced speed before a new or reworked mould goes to full rate. Skip cycles to save ten minutes of changeover time and you're trading it for stuck ware and early mould wear a few weeks out.

In 2014, on a five-line plant outside Brisbane, I found night shift had quietly dropped to two cycles as house practice. Nobody wrote it down. It just became how Thursdays worked. Reject rate on that line ran a steady 1-2% baseline in normal running, and it was spiking to somewhere between 8-12% in the first 15-30 minutes after every changeover, cold-mould checks and seed carryover stacking on top of gob-to-mould thermal mismatch before the line settled. Cut the seasoning short and that window doesn't shrink. It stretches.

23 minutes. That's what it cost most Thursdays, and nobody logged it as lost time.

Not a mould problem. A discipline problem.

The 0600 handover misses the previous shift's swabbing and seasoning-cycle count 70% of the time on most lines I've walked, because it was never built into the paper handover sheet. It's tribal knowledge, which is a polite way of saying it lives in one operator's head and leaves the building when he does.

Neck rings and plungers drift on their own clock

Most plants track overall mould body temperature and call the job done. Neck ring and plunger temperature mismatch is the one that gets missed, and it shows up downstream as finish checks and off-centre threads rather than an obvious body crack, so it gets diagnosed as a blank-side problem when it's actually a component homogenisation issue that started back at the preheat oven.

• The mould setter installs and aligns the mould halves and neck rings, and owns the physical fit before anything runs.
• The gaffer, or hot-end mechanic, sets plunger and blank timing plus cooling wind, and owns the thermal balance across cavities.
• The IS-machine operator verifies gob weight and distribution and signs off first-article ware, and owns the call to run at rate.

Poor handoff between those three roles loses more changeover minutes than any single piece of equipment does. An OEM-affiliated consultancy will benchmark your preheat time against the machine-builder's spec sheet, which doesn't hold up once your furnace is running a high recycled-content batch with a shifting cullet ratio. A generic Lean/Six Sigma boutique will time-study the changeover as one internal-versus-external SMED event and miss the coordination gap between setter, gaffer and operator entirely, which is where the real minutes go missing on an old Emhart running mechanical cam timing just as much as on a newer servo-driven IS machine.

A mould that's fifteen minutes early out of the oven isn't a scheduling win. It's a defect you haven't found yet.

Cheap gas or a shrinking allocation, the mould doesn't care

In the US, O-I has idled multiple North American furnaces since 2023 under its Fit to Win restructuring, concentrating volume onto fewer furnaces, which means each remaining line now runs more SKUs and more job changes per week to cover the volume the idled furnaces used to carry, per O-I's own investor disclosures. Ardagh's North American glass-packaging unit went through its own balance-sheet and network restructuring through 2024 and 2025, with energy and changeover-related downtime cited among the pressures on plant economics. Every job change on a smaller fleet now carries more weight than it did five years ago, and under 40 CFR Part 63 Subpart NNN, the Glass Manufacturing Area Source NESHAP, US plants only get so many "dirty" start-up cycles before they're chewing into permitted emission events. A rushed, out-of-spec mould change costs more than the 15 minutes it looked like it saved.

But the fuel side cuts differently by region. Saudi Arabia's tiered industrial gas pricing keeps feedstock cost for GCC glass manufacturers down around $1.25-2.00/MMBtu, a fraction of European TTF-linked industrial gas, per the Kingdom's Ministry of Energy schedules. That's good news for the fuel bill and bad news for the case to fix preheat discipline, because when energy's cheap the payback pressure on mould-thermal efficiency projects goes soft too. I've audited GCC plants where preheat was run as a scheduling afterthought precisely because nobody was watching the gas line item the way a European plant under EU ETS Phase IV's shrinking free allocation is forced to.

Standard work only works if every crew runs the same one

This is what Lean Glass builds the Job Change Tool around: one locked preheat curve, one seasoning-cycle count, one component-level temperature check, visible to every crew in real time instead of living in a notebook or a setter's memory. A vendor-neutral container glass consultant treats the mould setter, the gaffer and the operator as one coordination problem, not three separate time studies, because that's where the actual minutes disappear.

Lean Glass founder Zaid Hassoneh ran hot ends for the better part of two decades before this, starting on the floor at O-I Brisbane in 2005 and making plant manager by 2019, and won O-I's Most Improved Job Change Plant Globally award in 2016 for cutting exactly this kind of cross-crew variance out of a real plant floor, not a spreadsheet.

It's not glamorous work. It's the work. Go back to that 4am mould change on section six. Lock the preheat curve, hold the seasoning cycles at three to five no matter what the schedule board says, and give the setter, the gaffer and the operator a real handoff instead of an assumed one, and that Tuesday night shift starts looking exactly like the Wednesday day shift. That's the whole point of standard work: boring, on repeat, section after section.