What overheating assessors actually need...
What overheating assessors actually need from a noise report — and why vague briefs cost money Noise reports that...
WHY YOUR WINDOW SPEC JUST GOT HARDER — THE PART L 2026 U-VALUE CHANGES EXPLAINED
Part L 2026 has landed, and the headline window target looks reassuringly familiar — 1.2 W/m²K for windows with more than 60% glass, unchanged from the notional spec in Part L 2021. This means you can also consider windows with a higher U value, but you will need to compensate elsewhere in the SAP calculation. But the methodology for calculating that U-value has changed substantially, and the practical effect is that a window currently achieving a declared U-value of around 1.30 W/m²K may, under the new approach, calculate at closer to 1.40–1.50 W/m²K once the correct methodology and additional components are included. For projects with detailed or decorative window designs — Georgian bars, aluminium frames, or complex multi-light configurations — triple glazing is likely to be required in more cases than before. This post sets out the three key problems the change creates, three practical steps to manage them, and one piece of genuinely good news: SAP 10.3 has resolved a long-standing conflict between energy assessment performance and overheating compliance by changing how solar glass is treated in the calculation.
The target looks the same — but the methodology has changed
Until now, window U-values for dwelling energy assessments could be calculated using a standard reference window: a fixed size (typically 1230mm x 1480mm), assessed in a standard configuration. That approach allowed suppliers to declare a single U-value that covered most applications. Under Part L 2026, it is no longer permitted for new dwellings in England.
From 24 March 2027 — when the 12-month lead-in period ends — all window U-value calculations for new homes must use the actual size and style of each window as installed. The hotbox test method has been removed as an option. Calculations must follow BS EN ISO 10077-1 or 10077-2, applied to each window’s real dimensions and full configuration.
The notional U-value target of 1.2 W/m²K has not changed. Arriving at a calculated value below that figure is now harder, because the calculation must include more physical components — and most of those components perform worse thermally than the glazed unit itself.
The industry is currently developing the tools needed to run these calculations properly. They are expected to become available over the coming months, and in practice most of this work will sit with window manufacturers and suppliers, who will need to assess specifications correctly when pricing. The important message for design teams now: a stated U-value of 1.30 W/m²K from a supplier’s current product data should no longer be assumed to satisfy Part L 2026. It needs to be confirmed using the correct calculation method.
The new rules require frame sections, sills, transoms, dummy sashes and Georgian bars to all be included in the U-value calculation.
Three key problems
Problem 1: The same window now calculates at a higher U-value
The new calculation rules require the following elements to be included where present:
- Frame sections, sills and thresholds
- Add-ons and frame extenders
- Dummy sashes and transoms / mullions
- Decorative glazing bars — including Georgian bars and back-to-back spacers
- Non-dead load-bearing couplers (PVC H-couplers, aluminium wind load couplers)
Only dead load-bearing structural components — such as structural bay poles and two-part structural couplers — are excluded.
Each of these elements has a worse thermal performance than the glass unit. As more of them appear in the window configuration, the calculated whole-window U-value rises. The Glass & Glazing Federation demonstrated this compound effect at their May 2026 seminar:
| Configuration | Calculated U-value |
|---|---|
| Standard window | 1.27 W/m²K |
| With dummy sash | 1.27 W/m²K (minimal change) |
| With dummy transoms | 1.32 W/m²K |
| With PVC Georgian bars | 1.45 W/m²K |
| With aluminium Georgian bars | 1.60 W/m²K |
Starting from a window that currently declares at around 1.3 W/m²K, adding aluminium Georgian bars pushes the calculated value to 1.6 W/m²K under the new methodology. The window itself has not changed. The calculation has.
There is also an important point about roles here. A SAP assessor can set the required U-value target for a dwelling — for example, specifying that windows must achieve 1.30 W/m²K to satisfy the energy assessment. But they cannot calculate or certify the window U-value on behalf of the supplier. That calculation is the supplier’s responsibility, and it must use the correct Future Homes Standard methodology. Getting confirmation of this in writing, before the specification is fixed, is essential.
Problem 2: Every window now has its own U-value
Under the old approach, a supplier could provide a single declared U-value — calculated on the reference size — and that value could be applied across a range of products. Under Part L 2026, the U-value calculation must use the actual window dimensions.
This matters because U-value is not constant across sizes. Smaller windows have a higher frame-to-glass area ratio, which pushes the whole-window U-value up — frames perform worse thermally than double or triple glazed units. A window that achieves 1.2 W/m²K at 1200mm x 1400mm may fail at 800mm x 900mm.
The practical implication: a single U-value certificate no longer covers a product range. Each window — sized and configured as it will actually be installed — requires its own calculation. The GGF has described this as “almost a doubling of the work” for suppliers.
For SAP 10.3 (the interim compliance route running alongside the Home Energy Model), U-value inputs must reflect each window’s actual as-designed configuration. The days of entering a single declared value for all windows on a scheme are over.
Problem 3: Three types of window are most at risk
The methodology change does not affect all window types equally. Three situations are most likely to create compliance problems:
Georgian-style windows with glazing bars. As the GGF’s worked example shows, adding aluminium Georgian bars can push a window from 1.27 to 1.60 W/m²K — well above both the notional target and the 1.6 W/m²K England backstop limit. PVC bars perform better, but still carry a meaningful thermal penalty. Any scheme with Georgian-bar styling needs a proper calculation before the specification is confirmed.
Aluminium window frames. Aluminium already tends to perform worse thermally than uPVC or timber, and the new calculation methodology — which includes a fuller accounting of the frame — compounds this. Aluminium-framed windows are more likely to require triple glazing under Part L 2026 than their uPVC or timber equivalents.
Larger areas of glazing. Where glazing takes up a significant proportion of the external wall — large bifolds, floor-to-ceiling glazing, bay windows — even a modest increase in the calculated U-value has a proportionally larger effect on the dwelling’s overall energy performance. A window that was marginally compliant under the old method may pull the whole-dwelling calculation into failure when its true U-value is assessed.
Three key considerations
Consideration 1: Know what to ask your window supplier — and ask it early
The most practical step any design team or developer can take now is to ensure window suppliers are asked to confirm four things, in writing, before the specification is fixed:
- The calculated whole-window or whole-door U-value for each different opening size (values will vary across the schedule).
- The methodology used — it should reference BS EN ISO 10077-1 or 10077-2 and the Part L 2026 requirements.
- Whether the value is suitable for Future Homes Standard / Part L 2026 compliance.
- Whether glazing bars, frame type, door panels, spacers, and other components have been included in the calculation.
This is especially important where glazing specifications are fixed early for planning, design, or aesthetic reasons. A window schedule locked in at planning stage on the basis of current product data may not perform as expected when the correct Part L 2026 calculation is run. The earlier this is checked, the more options remain available.
The industry is developing integrated calculation tools that will eventually make this process more straightforward. Until those are widely available, the onus is on suppliers to confirm compliance using the new methodology. Design teams should treat any U-value figure that does not come with confirmation of the methodology and component inclusions as unverified.
Consideration 2: Review decorative features before they become a compliance problem
The window schedule for any scheme targeting Part L 2026 compliance should be reviewed for Georgian bars, dummy sashes, transoms, and frame extenders. These are no longer cosmetic decisions — they affect the U-value calculation directly.
Where Georgian bars are specified, material choice has a significant effect. PVC glazing bars perform better thermally than aluminium in the calculation. Where the bar material can be changed without affecting the design intent, this is the simplest fix.
Where dummy sashes or transoms are specified purely for appearance, it is worth asking whether they can be removed. If the visual effect is important to the client, replacing a dummy sash with a true opening light adds cost but removes the thermal penalty.
Sill depth also has a small effect. Stub sills (85–150mm) slightly worsen the calculation; larger sills (180mm+) provide a marginal improvement due to additional chambers in the section.
Consideration 3: Understand the limits of compensating through fabric
Where window U-values cannot be improved — because the design is fixed, the glazing area is large, or the frame type is constrained — improving the wider building fabric can help compensate in the SAP or HEM calculation. Wider cavity walls, full-fill PIR cavity insulation, aircrete blockwork, Hi-therm lintels, and additional thermal bridge reductions can all improve the overall dwelling energy performance and provide headroom to absorb a worse-performing window.
These measures are worth considering, but they have practical limits. On a scheme where multiple window types are affected, the cumulative impact on the energy assessment can be significant, and fabric improvements alone may not be enough to recover compliance. The earlier the window U-values are confirmed, the more time there is to assess whether fabric compensation is sufficient or whether the window specification needs to change.
For schemes where some windows cannot be simplified — bay windows with aluminium glazing bars on a traditional-style scheme, for example — triple glazing should be considered early in the cost plan. Triple glazing typically achieves centre-pane U-values of 0.5–0.7 W/m²K, which provides considerably more headroom to absorb the thermal penalty of complex frame configurations. The shift is not “all windows must be triple glazed” — many standard configurations will still comply with double glazing. But schemes with decorative detailing, aluminium frames, or a high proportion of smaller windows should carry a contingency for triple glazing on some units.
One piece of good news: the g-value change in SAP 10.3
Not everything in the new methodology works against you. SAP 10.3 has resolved a problem that caused difficulty on many schemes under the current Part L 2021 regime — the treatment of solar glass in the energy calculation.
Every window has a g-value (sometimes called solar factor or total solar energy transmittance) — a figure between 0 and 1 that describes how much solar energy passes through the glass. A g-value of 0.63 means 63% of solar energy is transmitted; 0.45 means only 45% gets through.
In SAP 10.2, the notional building — the reference against which a proposed dwelling is compared — was set with a glazing g-value of 0.63. If the proposed window had a higher g-value than 0.63 (say 0.66), that was advantageous for the SAP calculation: more solar gain meant more free heat, reducing the dwelling’s calculated energy demand. But if the proposed g-value was lower than 0.63, the opposite applied — the dwelling was penalised because it admitted less solar energy than the notional building assumed.
This created a direct conflict with overheating compliance under Part O. To reduce the risk of summertime overheating — particularly on south-facing elevations or rooms with large glazed doors — architects and assessors would often specify solar control glass with a low g-value. A large south-facing sliding door might be specified with a g-value of 0.45 to limit solar gain and help pass the overheating assessment. But in SAP 10.2, that same glass penalised the energy calculation, because the notional building expected 0.63 and was receiving 0.45 instead.
SAP 10.3 has resolved this. The notional building’s g-value now matches the proposed window’s g-value — whatever solar glass is specified, the SAP calculation does not penalise you for it. If a project needs solar control glass with a g-value of 0.45 to satisfy Part O or a TM59 overheating assessment, that value can be used without it pulling the energy calculation in the wrong direction.
This is a meaningful improvement for schemes where overheating is a genuine design constraint — south-facing glazing, large bifold or sliding door systems, properties in urban locations or weather files where cooling is a risk. It removes a trade-off that previously forced difficult choices between SAP performance and summer comfort. The two calculations are no longer pulling against each other on glazing specification.
What about SAP 10.3 and HEM — which calculation route applies?
SAP 10.3 is available immediately as an interim compliance route for Part L 2026. The Home Energy Model (HEM) — the longer-term replacement for SAP — has been delayed by at least three months from its original March 2026 publication date. Both systems will run side by side for at least two years after HEM launches, with six months’ notice before SAP 10.3 is phased out.
The U-value calculation methodology changes apply under both routes. Whether you are using SAP 10.3 or HEM, window inputs must be based on actual size and configuration, with all required components included. HEM goes further in requiring free opening areas for each sash and individual window-by-window inputs — but the core change to how U-values are calculated is the same in both systems.
One important implication: a window previously declared at 1.3 W/m²K on the standard reference size may calculate at 1.4–1.5 W/m²K or higher under the new methodology. If that window is entered into SAP 10.3 or HEM at its actual calculated value, the dwelling’s energy performance deteriorates compared to what was assumed at design stage. For schemes currently in design, this is worth checking now, before the specification is fixed.
Frequently asked questions
Does the notional U-value of 1.2 W/m²K still apply under Part L 2026?
Yes. The notional target for windows with more than 60% glazed area remains 1.2 W/m²K, and the notional target for doors and fire doors with less than 60% glazing remains 1.0 W/m²K. What has changed is how the U-value is calculated — using actual window dimensions and configuration rather than a standard reference size. The target is the same; achieving it is harder for many window types. So you will still be able to potentially relax this window U value target (to 1.30 W.n2,K or similar), but you need to balance this with improved U values and thermal bridging in the specification and it will heavily depend on the window size, design & components. Some windows may need to be triple glazed to achieve the same result.
Can I still use double glazing on a new build under Part L 2026?
Yes — but it depends on the window. Many standard configurations without decorative features will still achieve 1.2 W/m²K with double glazing. Windows with Georgian bars, aluminium frames, or small dimensions are more likely to require triple glazing under the new methodology. The only way to know is a calculation run on each window at its actual size and configuration. Similar to part L1 2021, the U value can be higher than the Notional Building, but you need to compensate elsewhere, often with Thermal Bridging.
Who is responsible for calculating the window U-value — the SAP assessor or the supplier?
The window supplier. A SAP assessor can set the required U-value target for the dwelling — for example, specifying that windows must achieve 1.30 W/m²K. But they cannot calculate or certify the U-value on the supplier’s behalf. The supplier must confirm that their product achieves the target using the correct Part L 2026 methodology, and that all required components have been included.
What has changed about how solar glass (g-value) is treated in SAP 10.3?
In SAP 10.2, the notional building was set with a glazing g-value of 0.63. If you specified solar control glass with a lower g-value — to reduce overheating risk under Part O — the SAP calculation penalised you for it. SAP 10.3 removes this problem: the notional g-value now matches the proposed window, so specifying solar control glass no longer affects your SAP score. This is good news for schemes where overheating is a design consideration.
What are the key transition dates?
For non-high-risk buildings: the 12-month lead-in period ends 24 March 2027. Building applications submitted and registered before that date can use the previous regulations, provided work starts before 28 March 2028. Transitions apply on a plot-by-plot basis, not by phase. For high-risk buildings, the lead-in extends to 24 September 2027.
Where can I get further guidance on which components to include?
The Glass & Glazing Federation is producing a support document covering inclusions and exclusions, with diagrams and cross-sections for PVC, aluminium, and timber frames. GGF technical documents 2.2 and 2.3 are also being updated, expected by end of 2026 / early 2027.
How Low Carbon Box can help
If you are working on a scheme that falls under Part L 2026 — or one where the window schedule is being finalised now — it is worth running a SAP 10.3 calculation before the specification is fixed. That way, any windows likely to fail under the new methodology can be identified and resolved at design stage, not at compliance stage.
Low Carbon Box carries out SAP 10.3 energy assessments for new dwellings across England and Wales, working with architects, developers and self-builders from early design through to as-built EPC lodgement. If you have a project where the window specification needs checking against Part L 2026 — or where overheating risk and solar glass specification need to be balanced — get in touch.
Find out more about our SAP energy calculations →
Low Carbon Box are accredited On-Construction Domestic Energy Assessors (OCDEA) providing SAP energy assessments and Part L compliance advice for homeowners, architects and contractors across the UK. Assessor reference EES/011387.
For more information or to discuss your project: info@lowcarbonbox.co.uk | 01858 322011
