Producing one kilogram of float glass generates around 0.9 kg of CO₂ — a figure surprisingly comparable to that of primary blast-furnace steel. The reason is the same in both cases: bringing a raw material to very high temperatures (for glass, over 1,500°C to melt the mix of silica sand, soda and limestone) requires an amount of energy that, today, is supplied almost everywhere by burning natural gas. But glass has a plant-level peculiarity that makes its decarbonisation even more complex than that of many other materials: a float-glass melting furnace, once lit, doesn't get switched off — it runs continuously for ten to fifteen years, day and night, because the thermal shock of shutting down and restarting would irreparably damage the furnace's refractory structure. This means every technological choice made when a furnace is fired up stays binding for over a decade.
Why a glass furnace can't simply be electrified overnight
Fully electrifying glass melting — replacing gas burners with immersed electrodes that heat the molten glass bath via the Joule effect — is technically mature and already used in small-to-medium-scale plants, especially for specialty glass and part of hollow-glass (container) production. The limit is scale: float-glass furnaces for construction are among the largest in the glass industry, melting hundreds of tonnes a day, and the electrical power needed for full electrification at this scale requires grid infrastructure that many European production sites don't yet have. The most common intermediate solution today is the hybrid furnace: part of the melting heat (typically 20–80% depending on configuration) is supplied by electrodes, while the rest still comes from fuel burners — a compromise that cuts emissions in proportion to the electric share, without requiring a full replacement of the existing plant.
Oxy-fuel combustion: burning better, not burning less
A second path, complementary to electrification, is oxy-fuel combustion: replacing air (78% inert nitrogen, which in conventional combustion absorbs a significant share of the burner's energy without contributing to melting) with pure or near-pure oxygen. Burning fuel in oxygen instead of air reduces the volume of combustion flue gas and noticeably improves the furnace's thermal efficiency, with a reduction in fuel consumption per tonne of molten glass estimated at around 10–20% compared with conventional air combustion. It's not a zero-emissions technology — the fuel burned is still usually natural gas — but it's an incremental measure that many glassworks have already adopted, because it requires relatively modest changes to the existing combustion plant compared with full electrification.
Cullet: the simplest lever, but a limited one
The most immediate decarbonisation lever, and already widely exploited, is increasing the share of glass cullet in the furnace batch relative to virgin raw materials: cullet melts at a lower temperature than the still-unreacted mix of sand, soda and limestone, cutting overall energy demand — as a rough guide, every 10% of cullet added to the batch cuts energy consumption by 2–3%. The problem, common to the entire flat-glass supply chain, is that the post-consumer float-glass cullet available to feed back into production is scarce: separate collection of glass from building demolition in Europe recovers less than 15% of the potential, against much higher rates for packaging glass. This means the room for improvement through cullet is today constrained not by melting technology but by the availability of sufficiently high-quality recycled material — a collection supply-chain problem, not a furnace one.
The overall picture suggests that decarbonising float glass will proceed through a sum of incremental measures — more oxygen, more electric power, more cullet — rather than through a single technological leap, precisely because of the plant-level constraint that locks each furnace in as an investment for a decade or more.