Glass, in theory, is one of the most recyclable materials there is: it can be melted and re-melted endlessly with no significant loss of quality, unlike many plastics that degrade with each cycle. And it's exactly this property that makes packaging glass — bottles, jars, containers — one of the most solid successes of European separate waste collection: in the EU the collection-for-recycling rate exceeds 80% (80.8% in 2023), against a global average of around 32%. Flat glass for construction, chemically almost identical, stalls instead at around 11% globally. The paradox has a precise explanation, and it has nothing to do with the chemistry of the glass itself.
Why cullet matters: energy and carbon dioxide
To understand what is lost when the glass from a façade ends up in landfill, you have to look inside the furnace. Float glass is made by melting, at around 1500 °C, a mixture of silica sand, soda ash, limestone and dolomite, to which cullet is added: glass that has already been melted once and then crushed. Cullet brings two benefits. The first concerns energy: it melts at a lower temperature than the raw materials and does not require the chemical reactions needed to turn them into glass. According to industry figures, every additional 10% of cullet in the batch cuts the furnace's energy consumption by about 3%. The second benefit concerns process emissions: soda ash, limestone and dolomite release carbon dioxide when heated, regardless of the fuel used. Cullet has already released those emissions the first time round, so it does not produce them again.
In addition, every tonne of cullet replaces about 1.2 tonnes of virgin raw materials, meaning sand that has to be quarried, transported and washed. Sand suitable for float glass is not as unlimited a resource as one might think: it must have a very low iron content and a regular grain size, and deposits with these characteristics are relatively few. Cullet is therefore not simply a way of disposing of waste: it is a high-quality raw material, already refined, which the building sector today largely throws away.
The problem isn't the glass: it's everything added to it
A packaging glass bottle is, in the vast majority of cases, pure soda-lime glass, with no laminates, no functional coatings, no bonded components. A facade glass pane, by contrast, almost always reaches end of life as a complex system: glass laminated with polymeric interlayers (PVB or SGP) that need to be mechanically separated from the glass before re-melting, Low-E coatings based on metal oxides (silver, titanium, tin) that, if not removed, contaminate the melting furnace and degrade the optical quality of glass made with that cullet, silicone sealants at the edges of the glazing unit, metal frames the glass is bonded to with structural adhesives. Each of these components requires a dedicated separation stage before the glass can go back into a furnace for new flat-glass production — and every separation stage has a cost that often exceeds the value of the recovered cullet.
Two very different streams of cullet must, however, be distinguished. Pre-consumer cullet arises before the glass reaches the building site: the edges trimmed on the float line, the offcuts from companies producing laminated glass and insulating glass units. It is clean, of known composition, collected at a few points, and indeed almost all of it goes back into the furnaces. Post-consumer cullet comes out of demolished or refurbished buildings: panes of different ages and manufacturers, often broken, mixed with frames, sealants, putty and debris. It is this second stream that is lost today. When it does not end up in landfill, it goes to lower-value uses: glass wool for insulation, container glass for bottles, road sub-bases, abrasives, foam glass. These are useful applications, but from the point of view of flat glass they represent an exit from the cycle: a window pane that has become a bottle will never become a window again.
Quality requirements: why float glass is more demanding than bottle glass
A second factor, less visible but just as decisive, is tolerance to impurities. Packaging glass accepts a relatively high share of cullet from mixed sources, because the optical properties required of a bottle are modest — it has to hold a liquid, it doesn't need to be perfectly clear and free of inclusions. Float glass for facades and windows has far stricter optical requirements: even minimal impurities in the cullet (fragments of ceramic, metal, borosilicate glass mixed in by mistake) can produce visible defects or structural weak points in the finished pane, making the entire production batch unusable for quality architectural applications. This means float-glass producers are willing to pay less for flat-glass cullet than packaging-glass producers pay for their own cullet, simply because they require a purity that's harder to guarantee — an economic disincentive that adds to, rather than replaces, the technical difficulty of separating composite materials.
Possible directions: design for disassembly and dedicated supply chains
The solutions under study in the sector move on two parallel fronts. The first is design-based: glazing-unit and facade systems conceived from the outset to be disassembled without destroying their components — reversible mechanical seals instead of permanent structural adhesives, laminated interlayers more easily separated from the glass through low-impact mechanical or chemical processes. The second front is infrastructural: creating collection supply chains dedicated specifically to flat glass from building demolition, separate from packaging-glass collection streams, with pre-treatment plants capable of efficiently removing laminates and coatings before re-melting. Neither direction alone would solve the problem: without a dedicated collection chain, even glass designed to be easily disassembled finds no industrial destination; without easily separable components, even the best collection network would gather a material too costly to purify to be competitive with virgin silica sand.
Before recycling: reuse and remanufacture
Remelting glass is the most visible solution, but not the best one. The circular economy hierarchy puts direct reuse first, then remanufacturing, and only then recycling, because every step down consumes more energy and destroys more of the work already done. In flat glass the most interesting case is that of insulating glass units. An insulating glass unit typically lasts twenty-five to thirty years, and it almost always fails not because the glass deteriorates but because the edge seal gives way: moisture gets in, the insulating gas escapes, condensation appears between the panes. The glass itself is still intact. Some pilot projects, mostly in northern Europe, have shown that units can be dismantled, the panes separated, cleaned and reassembled into new units, with updated performance thanks to a new spacer and a new gas fill.
The limit, for now, is practical. The panes of a 1990s building have sizes, thicknesses and coatings that differ from those required today, and toughened glass cannot be cut down. Reuse works well when the recovered panes are numerous and similar, as in the curtain walls of an office building, and much less well for scattered windows of different sizes. Laminated glass has its own route too: plants now exist that can crush the panes and separate the glass from the PVB interlayer, which in turn can be recovered and reused as a raw material for other products. These processes are still uncommon, but they show that the problem described above is technical and economic, not a law of nature.
The paradox of energy retrofits
One aspect makes the issue urgent right now. European and national policies for building energy efficiency are driving the replacement of millions of windows: single glazing and insulating units from the 1970s and 1980s are being replaced by low-emissivity double and triple glazing. It is the right choice in terms of energy use, because a good window considerably reduces heat loss. But every replaced window produces glass waste which, in most cases, ends up in the site skip along with the frame, the gaskets and the rubble. In other words, the same policy that cuts emissions during the building's use creates a large stream of glass that the system is not organised to recover.
Old glass, moreover, is often the easiest to recycle: a 1970s single pane has no low-emissivity coating or plastic interlayer, and is very close to a composition a float furnace can accept. The problem is purely logistical: separating it on site, keeping it clean, collecting it in sufficient quantities and taking it to a treatment plant. For anyone managing large-scale window replacement, such as a whole apartment block or a public building, providing a dedicated glass container and an agreement with a recovery plant is one of the actions with the best ratio of effort to result. It is also a good example of how the circular economy, in glass, depends less on technology and more on organisation.
What designers can do today
The first lever is documentation. A pane whose composition is unknown has little value as cullet, because no float manufacturer will take the risk of putting it into the furnace. Recording the type of glass, the manufacturer, the coatings and the interlayers in a materials passport or in the building's documentation turns, thirty years from now, an anonymous waste product into a traceable raw material. The second lever is the glazing detail. Mechanically fixed systems, with gaskets and pressure plates, allow panes to be removed whole; glass bonded with structural silicone is much harder to recover without breaking it. Where structural bonding is essential, it is worth planning from the design stage how and from which side the pane will be removed.
The third lever concerns the demolition site and procurement specifications. In the European waste classification, glass from construction and demolition has its own code, 17 02 02, and can be collected separately if it is removed before the rest of the building is demolished; mixed in with the rubble, it is lost. In Italy the Minimum Environmental Criteria (CAM) for construction, mandatory in public procurement, already require that at least 70% by weight of non-hazardous demolition waste be sent for recovery, and separated glass helps to reach that threshold. On the procurement side, glass manufacturers' environmental product declarations (EPDs) report recycled content and emissions per square metre: requesting and comparing them in tenders is the simplest way to reward those who invest in post-consumer cullet.
None of these choices requires technology that does not exist. They do, however, require thinking about the end of life of the glass at the design stage, something current practice almost never does. A façade designed today will stay in place for thirty or forty years: decisions about how the panes are fixed and what information is kept will determine whether, at that point, its glass goes back into a float furnace or ends up under the rubble.