Recycled Aluminium in Buildings: Recycled Content, EPDs and How to Specify It

Recycled aluminium is not one product. The documents we read describe it as a mix of primary ingot, pre-consumer scrap and post-consumer scrap whose proportions, and whose treatment in the carbon calculation, differ from one manufacturer and one programme to the next. This profile covers aluminium extrusions (window, door, curtain-wall and facade profiles) and rolled sheet used in buildings. It does not cover cast components, foundry alloys or aluminium foil.

We read EPDs registered with the International EPD System and issued by UL Environment, manufacturers' web pages, an alloy datasheet, the ASI Chain of Custody Standard and its guidance, a coating specification, a European Commission decision, the International Aluminium Institute's (IAI) page on recycling energy, and the BSI catalogue page for the self-declared claims standard (not the standard). Each section says which. This profile includes no EPD from Novelis, Constellium, Schüco or Alcoa, because we opened none, so rolled facade panels are covered only by the Aluminum Association's industry average. For other materials in the same family see recycled glass surfaces and recycled plastic panels.

How is recycled content declared for aluminium?

Three terms recur. Primary aluminium is made by electrolysis (the Hall-Héroult process, according to the IAI);^[14] pre-consumer scrap is generated during manufacturing (Hydro calls it production scrap, and the Aluminum Association's EPDs list post-industrial scrap separately from internal run-around scrap); post-consumer scrap comes from products that have reached the end of their use. The Aluminum Association's EPDs state that their definitions of internal process (run-around), post-industrial and post-consumer scrap are consistent with "ISO 14021/25 (2006)" and UL Environment's related interpretations, and they leave internal run-around scrap out of the composition declaration; the BSI catalogue page for the current ISO 14021:2026 states that the standard defines selected terms used in self-declared claims, and we did not read the standard or its definitions.^[9][10][17]

The documents then count the scrap differently. In Hydro's Nenzing EPD the content table assigns 0 kg CO₂e per tonne to the 81 % post-consumer material and 607 kg CO₂e per tonne to the 8 % pre-consumer scrap, and the text says pre-consumer scrap carries an environmental burden through co-product allocation.^[1] Hydro's Iberia EPD gives pre-consumer scrap the burden of the average aluminium mix consumed in Europe, treats post-consumer scrap as burden-free except for transport, and prints a second set of results for the other approach, in which pre-consumer scrap is also burden-free.^[2] Hydro's recycled-aluminium web page gives the reasoning: the emissions of post-consumer scrap were already accounted for in its first life, while production scrap, which has not yet completed a first life, keeps the footprint of its original production.^[11] Profal's EPD uses the ecoinvent "cut-off by classification" model, under which, it says, post-consumer aluminium has no impact as a raw material.^[5] Tecnoal's EPD prints a post-consumer share of 13.2 % in its content table, but in its module D calculation it sets the recycled input (MMR in) at 15 % on the stated basis that no post-consumer recycled input is used, and its resource-use table prints 0.371 kg of secondary material per kg; we show all three without deciding.^[6] ALsistem's EPD prints 2.36 % post-consumer recycled material against the aluminium (0.93 kg) of its worst-case product, in a column headed mass-% of product; elsewhere it prints 0.758 kg of secondary material per kg for that product and a recycled content of 75.8 % to 99.6 % across the included products, based on its suppliers' EPDs, without saying how these figures relate.^[7]

Hydro's two Iberia results show how large the modelling choice can be for one product: 5.60 and 2.99 kg CO₂e per kg of mill-finish profile for A1–A3, a difference of 2.61 (our subtraction), from the same EPD.^[2] Hydro's Nenzing and Ghlin EPDs say the zero-burden result is available from the company on request.^[1][3]

Brands, verification and mass balance

Hydro CIRCAL is described on Hydro's pages as recycled aluminium containing at least 75 % post-consumer scrap, with the share tracked for each production batch and the process verified by DNV; Hydro's CIRCAL 100R is described as made entirely from post-consumer scrap.^[12] The footprint is stated differently on different Hydro pages: an average of 1.9 kg CO₂e per kg on the global recycled-aluminium page; 1.9 kg CO₂e per kg in the EU and 1.5 in the US, marked as product and market dependent, on the global CIRCAL page; 2.3 kg CO₂e per kg or less on the US page about extrusions made with CIRCAL, updated 16 January 2026.^[11][12][13] None of the three pages states the life-cycle modules or the specific document behind its figure; we show all three and decide none. In the EPDs we opened, Hydro's CIRCAL 75R billet from Ghlin has a worst-case composition of 75.0 % post-consumer scrap, 8.01 % pre-consumer scrap and 16.84 % primary ingot, with 1.76 kg CO₂e per kg of billet for A1–A3, and the "Recycled Low-Carbon 75R" billet from Sjunnen (a forthcoming product, per its EPD) has 75.00 %, 16.30 % and 7.15 % with 1.28 kg CO₂e per kg of billet. Both are billet, not profile, and are declared per kg of billet.^[3][4]

The Aluminium Stewardship Initiative (ASI) Chain of Custody Standard (version 2, May 2022; the PDF also prints version 2.0.1) specifies two starting points for ASI aluminium, primary and recycled, connected by a mass balance system. It limits "Eligible Scrap" to pre-consumer scrap that meets traceability conditions, scrap assessed as post-consumer subject to supplier due diligence, and aluminium recovered from dross. In mass balance, output cannot be allocated as "partially CoC": if 20 % of an entity's outflow is designated CoC, that 20 % is 100 % CoC, as the standard puts it.^[15] Recycled content, with the methodology for pre-consumer and post-consumer scrap, appears in the standard only as optional sustainability data (criterion 9.3), and the guidance (version 2.1, April 2023) says that compliance with an optional criterion is not assessed for certification.^[15][16] Ask for the method behind any recycled-content figure.

How do the recycled-content and carbon declarations compare?

All results below are GWP-total for modules A1–A3 as printed, with the declared unit, modules, programme and PCR version as each document prints them. Declared units and system boundaries are kept in separate columns and nothing is converted between rows. The first table holds EPDs registered with the International EPD System under EN 15804+A2, each declared per 1 kg of profile; the second holds UL Environment EPDs under different category rules.

Source notes. Registry pages for the five International EPD System rows showed status valid (checked 12 October 2026); the Nenzing and Iberia PDFs print PCR versions 1.3.3 and 1.2.5 while the registry lists 1.3.4 for both, and the download link on each registry page points to the same PDF we read. We show both versions and do not resolve the difference.^{[1][2][5][6][7]} We did not check the UL SPOT listings for the Kawneer and Aluminum Association EPDs. Kawneer's and the association's figures are from different PCR families and primary-data years (2019–2020 and 2015–2019).^[8][9] For the Aluminum Association averages, 6.08E+03 kg CO₂e per tonne is 6.08 kg CO₂e per kg (our division, which the EPD says can be applied: results are converted to one kilogram by dividing by 1,000); the EPD's sensitivity analysis states that a one percentage point increase in primary aluminium raises global warming potential by as much as 117 kg CO₂e per 1,000 kg of semi-finished product.^[9] Packaging is handled differently: Hydro Nenzing's A1–A3 includes the balancing-out of biogenic carbon in packaging and ALsistem's EPD says packaging biogenic carbon is balanced in A3, while Profal declares A5, where its packaging end of life is modelled. The Nenzing and Profal EPDs both say using A1–A3 without module C is discouraged.^[1][5][7]

EPDs from different programmes, PCR versions and product types are not directly comparable. The International EPD System text printed in the Hydro and Profal documents (PCR 1.x) says two EPDs are comparable only if they use the same PCR (including the same version number) or fully aligned PCRs, cover products with identical functions and declared units, and have equivalent system boundaries, data quality, allocation and cut-off rules. The text printed in the Tecnoal and ALsistem documents (PCR 2.0.1) requires the same PCR including the same first-digit version number, identical declared units and identical scope in terms of included life-cycle stages.^{[1][2][5][6][7]} The rows above differ in all of these, and in whether a coating, anodising or thermal break is included. We rank no product and draw no conclusion about which is lower; a figure is meaningful for a project only against an EPD with the same declared unit, boundary and PCR.

How much energy does recycling save?

The IAI gives primary energy consumption for global primary aluminium production, from mine to cast house, as 186 GJ per tonne in 2019 and primary energy demand for recycled aluminium as 8.3 GJ per tonne, which it states as a saving of 95.5 % (8.3 ÷ 186 = 0.045, our division). It says there is no global estimate for scrap-based production, lists three regional datasets (Japan 5.7, Europe 8.5 in 2015 and 7.2 in 2018, North America 9.2 GJ per tonne), says the Japanese one covers wrought alloys only and was excluded from the average it used, and says its global primary figure is higher than the European (140) and North American (136) figures because of a higher share of coal.^[14] The Aluminum Association's EPD cover page puts the energy for recycled aluminium at 7 % of that for primary, and Hydro's page says recycling requires 95 % less energy; the three statements differ in basis and we do not reconcile them.^[9][11][14] The Aluminum Association's extrusion EPD adds that how primary aluminium is sourced matters: it models Canadian hydropower, Chinese coal-based, Middle Eastern gas-based and North American average supply as scenarios, and its Table 7 lists electricity carbon intensities for primary aluminium from 0.0421 kg CO₂e per kWh (Russia) to 0.778 (rest of the world).^[9]

When to specify it, and when not

Where the documents support it

• Window, door, curtain-wall and facade profiles. Profal's and Kawneer's EPDs name windows, doors and curtain walls among the applications of their profiles, Hydro Iberia's EPD names windows and doors, ALsistem's EPD covers profiles for windows and doors (its company description also lists curtain walls), and Hydro Nenzing's EPD names the building and construction sector; Tecnoal's EPD says extruded profiles are usually used for non-load-bearing structures such as door and window frames ("infixes" in its English text).^{[1][2][5][7][8][6]} These are manufacturer statements of intended use.
• Projects that need a documented recycled share and a carbon figure. The EPDs above print the metal input and a third-party verified GWP per declared unit. Our suggestion, not a document's: a request for tenders can ask for the post-consumer and pre-consumer share, the programme and PCR version, the modules, the declared unit and, where an EPD prints two results, which approach to pre-consumer scrap each follows.^[1][2][5]
• Same alloy specification as primary metal, as stated. The Aluminum Association's EPD says there is no relevant chemical composition difference between primary and secondary ingots where both follow the same alloy designation and composition limits, and Hydro states that CIRCAL meets the same technical and mechanical requirements as conventional aluminium alloys and is compatible with existing manufacturing and processing methods.^[9][12] Both are association or manufacturer statements; we read no independent test report.
• Thermal expansion, for design. Hydro's datasheet prints a coefficient of linear expansion of 23.5 × 10⁻⁶ K⁻¹ for EN AW-6063 (typical, at 20 °C).^[19] As an illustration (our calculation, free movement, no restraint): a 6 m length with a 50 K temperature change moves 6 × 50 × 23.5 × 10⁻⁶ = 0.00705 m, about 7 mm. We read no figure specific to recycled-content metal.

Where it needs care or may not suit

• A recycled percentage is not a carbon figure. Hydro Nenzing prints 81 % post-consumer and 8 % pre-consumer scrap with 2.00 kg CO₂e per kg, Hydro Iberia 31.8 % and 48.1 % with 5.60 (or 2.99), and Profal more than 99.5 % post-consumer with 1.14; the Nenzing and Iberia EPDs print their results under a stated approach to pre-consumer scrap.^[1][2][5] The documents do not attribute the differences between companies, and we do not.
• Corrosion and coatings. Hydro's alloy datasheet grades corrosion resistance 5 for 6060 and 6063 and 4 for 6005, 6005A and 6082 on its relative scale.^[19] For organic coatings, the QUALICOAT Specifications 2023 (applicable from 1 July 2023) define a SEASIDE endorsement for which the enhanced pretreatment must remove at least 2.0 g/m² of aluminium by etching, against at least 1.0 g/m² for the standard pretreatment; anodic pretreatment carries the SEASIDE endorsement automatically and also requires a minimum etching rate of 2 g/m², and a filiform-corrosion laboratory test is part of the SEASIDE inspection.^[20] We found no statement in the specification's text on the sites or atmospheres for which SEASIDE is intended; linking it to coastal projects is a reading of its name, not a statement in the specification. The specification also mentions a research programme, QUALICOAT 3.0, on alloying elements, extrusion quality, microstructure and corrosion potential of both primary and recycled aluminium, described in an appendix (A13) that we did not read. We found no document that compares the corrosion or coating performance of recycled-content and primary profiles.
• Anodising and colour. Hydro's datasheet rates alloys 6060 and 6063 5 for suitability for decorative anodising on its relative scale; it is an alloy datasheet and does not address scrap content.^[19] QUALICOAT requires the aluminium to be suitable for the coating processes, free from corrosion and free from contaminants.^[20] We found no document on colour consistency or anodising response of profiles with a high share of post-consumer scrap; our suggestion is to ask the extruder and finisher for samples from the intended billet supply.
• Thermal bridging in facades. Hydro's datasheet gives 201 W/(m·K) for 6063 T6 and the Hydro Iberia EPD 118 to 174 W/(m·°C) for 6xxx alloys; the documents print different numbers for different alloys, tempers and sources, and we do not reconcile them.^[19][2] ALsistem's EPD describes assembling profiles with plastic strips (thermal breaks) as improving thermal insulation by reducing heat and cold transfer; its content declaration lists the thermal break as 0 to 0.19 kg per kg of profile, in ABS or polyamide, with 0 to 96 % post-consumer content.^[7] We opened no frame U-value or thermal test report.
• Fire. Commission Decision 96/603/EC lists aluminium and aluminium alloys (not in finely divided form) among materials classified A1 and A1FL without testing, and says products coated with an inorganic layer may also be so classified; it also states that no listed material may contain more than 1.0 % homogeneously distributed organic material and excludes panel products with organic layers.^[18] The text we read does not address painted or powder-coated profiles beyond those conditions. Hydro's datasheet prints a melting range of 615 to 655 °C for 6063 and the Iberia EPD 495 to 640 °C for its 6xxx range, and we read no document on how aluminium members behave in a fire.^[19][2] We opened no fire-resistance test for an aluminium facade or panel system; the Decision covers the listed materials and products made only from them.
• End of life and "infinitely recyclable". Hydro's page says aluminium can be recycled infinitely without losing its qualities and the Aluminum Association's EPD cover page calls it infinitely recyclable; these are manufacturer and association statements.^[11][9] The EPDs model less than full recovery: 96 % collected and 95 % of that sent to recycling after shredding (Hydro Nenzing), 95 % recovery and 5 % landfill (Hydro Iberia), 85 % recycled and 15 % landfilled (Profal), 95 % (Tecnoal and the Aluminum Association), and 90 % (ALsistem).^{[1][2][5][6][9][7]} These are scenario assumptions, not measured rates. Module D credits depend on the scenario: Nenzing −1.42, Iberia −4.57, Tecnoal −4.51, ALsistem −0.279 for its worst case, and zero for Profal, whose EPD says a 100 % secondary input is not considered in module D.^{[1][2][6][7][5]}
• Supply and traceability. Hydro says recycling post-consumer aluminium requires advanced sorting, alloy control and process technology.^[12] A batch-level claim is only as good as the chain behind it: ASI's mass-balance standard lets recycled-content figures be optional data with a stated method,^[15][16] and we opened no ASI certificate, audit or registry entry for any supplier. Our suggestion: ask for the document that supports the claim for the billet actually delivered, not for an average.
• Comparison with other frame or cladding materials. We opened no document comparing aluminium with timber, PVC-U or steel, so this profile makes no such comparison.

Suppliers whose documents we read

Frequently asked questions about recycled aluminium in buildings

The bottom line on recycled aluminium

In the documents we read, the useful question about recycled aluminium is not how much recycled metal a product has but which scrap, counted how, in which document. The share of post-consumer scrap, the treatment of pre-consumer scrap, the declared unit, the PCR version and whether finishes and thermal breaks are included decide what an EPD figure means, and they differ between manufacturers and programmes. The Aluminum Association states that primary and secondary ingot under the same alloy designation do not differ in relevant chemical composition and Hydro states that CIRCAL meets the same requirements as conventional alloys; the documents differ on carbon figures and on end-of-life assumptions. We did not read the text of EN 13501-1, EN 15804, ISO 14021 or ISO 14025, we did not open the QUALICOAT 3.0 appendix, we opened no EPD from Schüco, Novelis or Constellium, and we read no independent test report comparing recycled-content and primary aluminium; the figures above are what the named documents state.

Related materials