Reusing a window instead of buying a new one saves roughly 58 to 63 % of its embodied carbon per square metre, according to the only published life-cycle study that has actually modelled window reuse: the Öko-Institut’s 2003 assessment for the Bauteilbörse Bremen, which found a new timber insulated-glazing window at 58.0 kg CO₂e/m² against 24.2 kg CO₂e/m² when reused, and a new PVC window at 120.0 against 44.4. Those figures are twenty years old and carry a specific assumption about residual life, so treat them as the order of magnitude, not the answer for your project. For a current new-window baseline, Glass for Europe’s 2022 review of EPDs puts a Low-E double-glazed window at 113.3 kg CO₂e/m².
That is what the public record offers. This article lays it out, says what the numbers assume, and explains why the reuse decision on windows is one of the few where operational energy can genuinely overturn the embodied-carbon case.
The short version
- Öko-Institut (2003, for Bauteilbörse Bremen): new timber IGU window 58.0 kg CO₂e/m², reused 24.2 → saving 33.8 kg CO₂e/m² (58 %). New PVC window 120.0, reused 44.4 → saving 75.6 (63 %).
- The study credits a reused window with half the residual service life of a new one (timber 20 vs 40 years, PVC 25 vs 50). Change that assumption and the percentage moves.
- Current new-window baseline: Low-E double glazing 113.3 kg CO₂e/m², triple 140.4, from EPDs of aluminium- and PVC-framed products (Glass for Europe, October 2022).
- Windows are the exception to “reuse always wins”: an old unit’s U-value can cost more operational carbon than its reuse saves. Check both.
- No current EPD or dataset exists for a reclaimed window. Elementa treats reuse as an indicative pathway factor (0.15 × new), not a certified value.
- All figures here are indicative, for early comparison, not for certification or compliance.
What does the Öko-Institut study actually say?
The study, Ermittlung der durch die Wiederverwendung von gebrauchten Bauteilen realisierbaren Energieeinsparpotenziale und CO₂-Reduktionspotenziale (Öko-Institut e.V., Freiburg, final report 1 December 2003, project lead Dr. Dietlinde Quack), was commissioned to quantify what a component-reuse exchange in Bremen was saving. It compared a reference scenario (a new component is manufactured; the old one is disposed of, with credits for material substitution and energy recovery) with an alternative scenario (the old component is removed, refurbished and reinstalled) for around twenty component types, windows among them (Öko-Institut, 2003, PDF, Annexes VII–X, pp. 45–48).
For windows, per m² of standard window:
| Window type | Reference (new) | Alternative (reused) | Absolute saving | Relative saving |
|---|---|---|---|---|
| Insulated-glazing window, timber frame | 58.00 kg CO₂e | 24.20 kg CO₂e | 33.80 kg CO₂e | 58.3 % |
| Insulated-glazing window, PVC frame | 120.04 kg CO₂e | 44.41 kg CO₂e | 75.63 kg CO₂e | 63.0 % |
Primary energy tracks the same pattern: 1,866 MJ/m² new timber window against 737 MJ/m² reused (60.5 % saving); 2,070 against 756 MJ/m² for PVC (63.5 %).
Two assumptions drive these numbers and both are stated in the report. First, the reference scenario gives the new window credits for what happens to the old one at end of life (material substitution, and electricity from incinerating the old timber or PVC), which is why the timber window’s reference figure is comparatively low: the report notes a “comparatively high credit in the case of the timber window”. Second, and more important, the reused window is credited with only half the residual service life of a new one. The study’s service-life table gives a timber insulated-glazing window 40 years new and 20 years reused, and a PVC window 50 and 25. The reuse figure is therefore the burden of half a window’s manufacture plus refurbishment and reinstallation, not zero.
Citation capsule: In the Öko-Institut’s 2003 study for the Bauteilbörse Bremen, reusing an insulated-glazing window instead of manufacturing a new one saves 33.8 kg CO₂e/m² for a timber-framed unit (58.0 → 24.2 kg CO₂e/m², 58 %) and 75.6 kg CO₂e/m² for a PVC-framed unit (120.0 → 44.4, 63 %), assuming the reused window delivers half the residual service life of a new one.
What does a new window cost in carbon today?
The 2003 figures predate EN 15804 and today’s heavier, better-performing units. For a current baseline, Glass for Europe’s October 2022 paper on the embodied and operational carbon of glazed windows took mean values from four EPDs (two aluminium-framed products, two PVC-framed) and arrived at (Glass for Europe, 2022, PDF):
| Window | Build-up | Uw | Weight | Embodied CO₂ |
|---|---|---|---|---|
| Low-E double glazed | 4-16-4, one Low-E coating, frame fraction 0.3 | 1.3 W/(m²K) | 39.5 kg/m² | 113.3 kg CO₂e/m² |
| Low-E triple glazed | 4-16-4-16-4, two Low-E coatings, frame fraction 0.3 | 1.0 W/(m²K) | 51.2 kg/m² | 140.4 kg CO₂e/m² |
Note the basis: an average across aluminium and PVC frames, not a timber window, and cradle-to-gate manufacturing emissions. A timber-framed EPD would generally sit lower on GWP-fossil. If you apply the 2003 study’s reuse ratio to this baseline you get an indicative reuse saving in the region of 65–70 kg CO₂e/m² for a double-glazed unit. That is arithmetic across two sources with different boundaries; use it to size the opportunity, not to report it.
Citation capsule: Glass for Europe’s 2022 review of manufacturer EPDs gives a mean embodied carbon of 113.3 kg CO₂e/m² for a Low-E double-glazed window (Uw 1.3, 39.5 kg/m²) and 140.4 kg CO₂e/m² for a Low-E triple-glazed window (Uw 1.0, 51.2 kg/m²), averaged across aluminium- and PVC-framed products.
Why are windows different from other reused elements?
For a brick wall or a steel beam, reuse almost always wins on whole-life carbon: the embodied saving is large and the element’s performance does not change the building’s energy use much. Windows break that pattern. The unit you reclaim has a U-value, and if it is a 1990s double-glazed unit at Uw ≈ 2.5–3.0 W/(m²K) replacing what would otherwise be a new unit at 1.0–1.3, the extra heat loss over the window’s life is operational carbon that has to be set against the embodied saving.
The Glass for Europe paper is a useful reminder of the scale. In its single-family-home model (162 m² floor, 32.4 m² of window), moving from Low-E double to Low-E triple glazing, a difference of 27 kg CO₂e/m² embodied, cuts operational emissions by 765–930 kg CO₂ over 30 years in central and northern climates. Divided over 32.4 m² that is roughly 24–29 kg CO₂ per m² of window: the same order as the embodied difference. A step from an old, uncoated unit to a modern one is several times larger.
So the honest reuse question for windows has two halves:
- Embodied: what does reuse save against new? Indicatively 55–65 % of the new unit’s manufacturing carbon, per the only study that has measured it.
- Operational: what does the reclaimed unit’s U-value cost over its remaining life, against the new alternative, in your climate and with your heating source? For a well-insulated envelope on a decarbonised grid, often little. For a leaky envelope on gas, potentially more than the embodied saving.
Reclaimed windows make most sense where the second half is small: internal partitions, unheated spaces, buildings where the reclaimed unit is already a decent double-glazed product, or projects where the old windows are retained in place and secondary glazing is added. For the general logic of when retention loses to replacement, see reuse vs new build.
Citation capsule: Window reuse is one of the few cases where operational carbon can overturn the embodied-carbon case for reuse, because a reclaimed unit’s U-value directly changes the building’s heating energy. Glass for Europe’s 2022 model shows a 27 kg CO₂e/m² embodied step between double and triple glazing being offset by 765–930 kg CO₂ of operational savings over 30 years across 32.4 m² of window in central and northern European climates.
How does Elementa handle a reclaimed window?
There is no EPD for a reclaimed window and no generic dataset for one in ÖKOBAUDAT or the ICE Database. Anything a tool shows you for a reused window is a modelling convention, and it should say so. Elementa’s convention is the reuse pathway factor: a reused element is assigned an indicative 0.15 × the embodied carbon of its new equivalent, alongside 0.05 for retained in place and 1.00 for new. That is deliberately in the same range as the Öko-Institut’s finding (a reused window at 42 % and 37 % of new, before the residual-life haircut), and it is labelled indicative because it is.
What Elementa does not do is net the operational side. Window U-values and heating energy sit in the operational assessment, which is a separate calculation. In pre-assessment the useful move is to flag the element: “reclaimed window, Uw unknown, operational trade-off to be checked”. That is more honest than either ignoring the trade-off or inventing a number for it.
Citation capsule: No EPD or generic dataset exists for reclaimed windows. Elementa represents window reuse with an indicative pathway factor of 0.15 × new (retain 0.05, new 1.00), consistent in scale with the Öko-Institut’s 2003 finding, and flags the operational U-value trade-off for separate assessment rather than netting it.
Frequently asked questions
How much embodied carbon does reusing a window save?
Indicatively 55–65 % of the new window’s manufacturing carbon. The Öko-Institut’s 2003 study for the Bauteilbörse Bremen found 58 % for a timber insulated-glazing window (33.8 kg CO₂e/m²) and 63 % for a PVC one (75.6 kg CO₂e/m²), crediting the reused unit with half the residual life of a new one. No newer study has measured window reuse directly.
What is the embodied carbon of a new double-glazed window?
Around 113 kg CO₂e/m² for a Low-E double-glazed unit and 140 kg CO₂e/m² for Low-E triple glazing, per Glass for Europe’s 2022 average of four manufacturer EPDs (aluminium and PVC frames). Timber-framed products generally sit lower on fossil emissions. Always check the specific product EPD.
Is a reclaimed window lower carbon over its whole life?
Not automatically. If the reclaimed unit has a much worse U-value than the new alternative, the extra heating energy over its remaining life can exceed the embodied saving, especially in a poorly insulated building heated by gas. Reuse is strongest where the unit is already a reasonable double-glazed product or where the window’s thermal performance matters little (partitions, unheated spaces).
Are there EPDs for reclaimed windows?
No. EPDs cover manufactured products. Reclaimed components have no EPD and no generic dataset, so any figure for them is a convention. State the convention when you use one.
Can Elementa certify a reuse saving?
No. Elementa gives indicative pathway factors and impact classes for early comparison. A certified figure needs a formal LCA by a qualified assessor, and for windows it needs the operational side calculated too.
The takeaway
One study, twenty years old, is the entire public evidence base for window reuse, and it says 58–63 % embodied saving on the assumption of half a service life. A new double-glazed window today carries about 113 kg CO₂e/m². Those two facts size the opportunity. What decides it is the reclaimed unit’s U-value against the alternative in your building, which is an operational calculation the embodied number cannot stand in for.
If you want reuse, retain and new compared per element, with the assumptions visible, before formal assessment, start a project.