What EU ETS Phase IV Actually Does to Container Glass Margins

The number most plant managers in European container glass have memorised is 1.74. That is the linear reduction factor from Phase III of the EU ETS. Under the Fit for 55 revision, that figure became 4.3% annually from 2024. If your operations team is still modelling carbon exposure against the old rate, the gap between projected free allocations and actual entitlement by 2027 will not be a rounding error. It will be a line item in the EBITDA bridge that nobody planned for.

This is not an abstract regulatory risk. European container glass plants are energy-intensive by nature, running furnaces above 1400°C around the clock. Every GJ per tonne of pull has a carbon cost attached. Under EU ETS Phase IV, that cost is moving on two fronts simultaneously. EUA prices are structurally higher than they were in 2018, and the volume of free allowances your plant receives is shrinking on a defined schedule. Both fronts move against margin.

The free allocation gap is larger than most plants have modelled

Container glass is still listed as a carbon leakage sector under EU ETS Phase IV, which means eligible plants receive free allowances at 100% of their benchmark. That sounds protective until you look at how fast the benchmark itself is being revised downward. The benchmark is expressed in tCO2 per tonne of glass produced. A plant running above that benchmark, whether from high specific energy consumption, a low cullet ratio, or aged furnace technology, is already buying allowances on the open market. Every tonne of CO2 above the benchmark is a purchased EUA.

In 2021 I was reviewing energy performance at a four-furnace plant in western Europe: two end-fired regenerative furnaces that had been running since the early 2000s, plus two newer cross-fired units commissioned in 2014. The plant had sailed through Phase III on comfortable allocations. When we mapped actual emissions against the Phase IV benchmark trajectory, the shortfall came to roughly €3.2M per year in purchased EUAs by 2025, at then-current prices. I asked the plant manager and the operations VP what their ETS exposure number was. Nothing. Not an estimate. Not even a range.

That is not unusual. Phase III was generous enough that most plants never built internal carbon accounting with any real teeth. The revised Linear Reduction Factor of 4.3% per year ends that era. A plant sitting exactly on its benchmark today will see free allocation erode at a rate that outpaces most energy efficiency investment cycles. A plant running above benchmark is already in the EUA market, whether it knows it or not.

Under Phase IV, monitoring and reporting requirements have also tightened. Plants are now expected to demonstrate compliance at a level of detail that requires credible internal tracking of specific energy consumption, cullet ratios, and production tonnes at the furnace level. The days of a compliance-only ETS posture are over. Filing once a year and forgetting about it until the allocation arrives is no longer a viable approach.

What the Market Stability Reserve is doing to EUA prices

The Market Stability Reserve was designed to do exactly what it is doing. Introduced in 2019, it absorbs surplus allowances when the total number in circulation exceeds defined thresholds, permanently removing them from the supply pool. The effect on EUA prices was immediate. Spot prices moved from below €10 per tonne in 2017 to above €80 by 2022. The MSR's intake rate was doubled through 2023 to accelerate the drain of the Phase II and Phase III overhang that had kept prices artificially suppressed for nearly a decade.

The structural direction of EUA prices through 2030 is up, moderated by economic cycles but not reversed by them. A container glass plant buying allowances on spot in 2028 is not pricing at €25 per tonne. Most European energy market agencies model the trajectory at €100–130 per tonne under Fit for 55 assumptions. A plant pulling 300 tonnes per day of flint container glass, running at 4.8 GJ per tonne with 35% cullet, carries a carbon intensity above the benchmark for that product category. At €100 per EUA, a 20,000-tonne annual shortfall against free allocation is €2M of direct cost (and yes, at €130 per tonne that number becomes €2.6M, and you won't be the only one in the room asking whether the furnace needs to come forward in the capital queue).

And the MSR absorption mechanism is not reversible under normal market conditions. That is the structural point for capital planning. You are not waiting for the carbon market to correct. You are managing a defined cost curve with a known direction.

Where container glass gets squeezed harder than other sectors

Container glass has a structural disadvantage in ETS that ceramic tile or float glass producers don't face to the same degree: the capital cycle mismatch. A container glass furnace runs 12-18 years between major rebuilds. If your furnace was rebuilt in 2019, you are living with that specific energy consumption figure until at least 2031-2033. Electric boosting can be retrofitted on most regenerative furnaces and typically delivers an 8-12% reduction in specific energy consumption depending on boost level and furnace geometry. A full oxy-fuel conversion or electro-melter is a furnace rebuild project, not a retrofit. Most European operators know that.

The other structural constraint is cullet. Higher cullet ratios directly reduce melting energy: roughly 2.5-3% per 10 percentage point increase in cullet content. For green and amber container glass, post-consumer cullet availability in northern and western Europe is reasonable. For flint glass it is not. Mixed-colour cullet cannot go into clear or lightly tinted products without introducing cords and colour non-uniformity that fail inline inspection. A flint container glass line running at 30% cullet because clean cullet supply is limited is paying a carbon penalty that no forming optimisation will fully offset.

The furnace rebuild decision is where ETS costs most directly influence capital planning. A plant facing a rebuild between 2025 and 2028 has a genuine choice between conventional regenerative, oxy-fuel, and hybrid electric configurations. The carbon cost differential over a 15-year campaign life, modelled at €100-130 per EUA, can add €15-25M to the total cost of ownership of a conventional rebuild compared with a higher-capital electric hybrid. That calculation was not seriously on the table in 2015. It is now.

The controllable levers before the next furnace rebuild are real but bounded. In practice they come down to four areas:

• Cullet ratio maximisation: target 60%+ for green container glass, constrained by colour spec and supply for flint.
• Electric boost retrofit: 8-12% specific energy reduction on most regenerative furnaces. Payback period improves significantly above €80 per EUA.
• Pull rate discipline: minimum viable pull at target pack efficiency reduces CO2 intensity per tonne. A 1% pull reduction at stable pack moves your ETS position.
• Forming efficiency: a gob weight CV at or below 0.4% ensures every melted tonne is packed. Every reject is carbon already spent.

The carbon cost per tonne of packed container glass is a hot-end number, not a finance number. The day the hot-end superintendent stops owning it, the plant starts losing ground on its ETS position every single shift.

What this means for your planning cycle before 2030

The CBAM, the Carbon Border Adjustment Mechanism, enters full implementation in 2026. It applies a carbon price to the embedded emissions of goods imported into the EU, including glass products. For container glass producers in the GCC or North Africa exporting into European markets, this changes the competitive calculation. But European producers should not treat CBAM as a margin protection tool. It levels the playing field on imports. It does nothing for your own cost position in the ETS market.

The plants that carry Phase IV through to 2030 with intact margins will be the ones that treat carbon cost as an operational variable, not a compliance cost. The hot-end superintendent should see the ETS position monthly alongside OEE. The forming efficiency programme should connect to the carbon account, not sit siloed in a quality dashboard. The next furnace design brief should carry a CO2 intensity target in tCO2 per tonne of packed glass, right alongside pull rate and fuel consumption.

Look, most plants I've walked into in the last three years have the Phase IV schedules somewhere in the building. What they don't have is the operational translation of that data into something an operator on the hot end can act on. That is the gap.

If your plant hasn't mapped its actual emissions against the Phase IV benchmark trajectory, that work needs to happen before the next furnace campaign planning cycle. A hot end audit gives you CO2 intensity per tonne across each section, benchmarked against the EU ETS Phase IV allocation schedule, with the exposure quantified in euros. That is the starting point for a defensible carbon plan. Lean Glass works as a vendor-neutral container glass consultant, independent of any OEM equipment roadmap. Our strategic advisory engagements with European plants start with that Phase IV exposure calculation and work backwards through the capital and operational decisions that close the gap. If you want an European container glass consultancy that will give you the numbers straight before 2030, that conversation is worth having now.