European Container Glass: Old Plants, Sharper Operators

In 2014 I walked a five-line plant outside Lyon during a cold repair delay that had already run four months over schedule. The furnace was on borrowed time, the mould shop was patching swing plates that should have been retired two campaigns earlier, and the plant manager kept telling head office the numbers were 'stable'. They weren't. Pack-to-melt was sitting at 81% and nobody upstream had asked why.

That plant wasn't unusual. It was Europe.

The fleet is old and that's the actual story

Container furnace campaigns run ten to fifteen years continuous before a cold repair, and a full rebuild costs on the order of USD 30-50 million. That single capital decision sets a plant's cost position for the next decade. It's the real strategic lever in this business, not another inspection retrofit or robotics upsell.

A lot of the fleet across France, Germany, Spain and Italy was built or last rebuilt well before the current energy and carbon cost stack existed. Best-in-class container furnaces run at roughly 3.5-4.5 GJ per tonne of glass melted. Plenty of legacy European furnaces sit measurably above that band, and every gigajoule above it now carries a carbon price attached. Rebuild timing has stopped being a maintenance calendar question. It's a competitiveness question.

And the gas shock of 2022-2023 made that brutally visible. Multiple continuous-melt furnaces across Ardagh Glass Europe and O-I Europe sites in Germany and the UK were curtailed during that period, because a container furnace can't just be cold-idled the way a discrete manufacturing line can. Shut it down wrong and you're not pausing production, you're funding an early rebuild.

ETS Phase IV is squeezing faster than efficiency gains can offset it

Under EU ETS Phase IV, free allocation for glass manufacturing sits on a declining benchmark curve, and the linear reduction factor — 4.3% annually, rising further under the revised Fit for 55 trajectory — is cutting free permits faster than most plants can improve output efficiency to keep pace (European Commission DG CLIMA). That gap shows up as a direct margin line, not a footnote in a sustainability report.

Glass packaging isn't inside CBAM's product scope yet — cement, steel, aluminium and fertilisers are, container glass isn't — but FEVE and Glass Alliance Europe have been lobbying hard for inclusion, arguing the carbon leakage risk from non-EU container imports is understated. Whether that lands in the next revision or not, any capital case built without pricing that possibility in is already out of date.

Nobody gets fired for missing a job change target. Plenty of European plants have been quietly bled dry by one.

Cullet is a carbon lever before it's a sustainability slide

FEVE reports EU average collection-for-recycling rates for glass packaging in the 76-80% range, with the voluntary 'Close the Glass Loop' commitment targeting 90% average recycled content by 2030 — and Belgium, Germany and Switzerland are already past that line. Every ten-point increase in cullet ratio cuts furnace energy consumption by roughly 2-3% and CO2 emissions by around 5%. That's not a marketing number. That's a furnace running cooler, a redox profile that's easier to hold, and a carbon bill that's smaller before you've touched a set point.

Verallia, Ardagh, O-I Europe and Vidrala carry the large majority of EU container tonnage between them, and the consolidation pressure on smaller independents is rising because they can't fund a furnace rebuild under the current energy and carbon cost stack on their own. Cullet supply security is going to decide which of the mid-sized independents are still running their own furnace in 2032 and which have been folded into somebody else's network.

Where a container glass consultant actually finds the margin

Here's what the OEM-affiliated consultancies keep missing. They walk in, see reject rates and a tired inspection line, and prescribe a capital upgrade — new Iris-type cold-end inspection, an IS machine retrofit, more automation. World-class plants run under 5,000 ppm total reject at cold-end inspection. Average operations run 10,000-15,000+ ppm or worse, and the gap is almost never the inspection machine. It's mould maintenance discipline (check the centring tool reading first, and yes, I know your fitter says it's fine, check it anyway).

Generic Lean and Six Sigma boutiques have a different blind spot. They bring a DMAIC framework built for a line you can stop and reset. A container furnace on a 24/7 campaign doesn't work that way. Melter thermal inertia constrains every downtime and containment decision, and a consultant who's never stood a night shift on an IS machine floor tends to miss that entirely.

The KPI that actually reflects hot-end health is pack-to-melt — good ware packed against glass melted — not the OEE dashboard everyone screenshots for the board. World-class runs above 90%. Many EU legacy furnaces sit at 80-85%, and job-change losses are usually the biggest single chunk of that gap. Mould-change competency under two hours per job change is the highest-variance human factor in defect rate on most European lines I've audited, and it doesn't show up on a generic OEE report at all.

On a cross-shift basis that variance typically runs 30-60% on identical SKUs in plants that haven't systemised the changeover — same recipe, same mould set, wildly different first-ware quality depending on who's running the 0600 handover. That's the gap our Job Change Tool was built against: a versioned SKU library, a live execution checklist mapped to the nine-stage Job Change Lifecycle, and KPI tracking that trends section-level variance instead of hiding it in a monthly average.

• Blisters trace back to entrapped gas, usually a batch or refractory issue, not a forming fault
• Cord points at chemical inhomogeneity in the melt, not a mould problem
• Checks and overpress on NNPB lines usually mean forehearth temperature has drifted past ±2°C
• Underfill and dogboning almost always trace to swing/blank timing mismatch at job change, not glass composition

A lot of that gets misdiagnosed because the person doing the diagnosis is selling something upstream of it. A vendor-neutral glass consultant doesn't have a machine to sell, which changes what they're willing to tell you about your own mould shop.

What this means for the next capital cycle

Europe's container glass market isn't shrinking because the product is wrong. It's under margin pressure because the cost stack — energy, carbon, ageing refractory — is stacking up faster than most plants have adjusted their operating discipline to match. The plants that come out ahead over the next rebuild cycle won't necessarily be the ones with the newest equipment. They'll be the ones who fixed job change variance and cullet sourcing before they spent the fifty million on the furnace.

We work across Europe container glass plants on exactly this gap, usually starting with a hot-end audit before any rebuild conversation happens. If your board is weighing a cold repair against a network consolidation decision, that's a strategic advisory conversation worth having before the capital paper gets written, not after.