What are thermal bridge psi values at window junctions and how do I get them right?

A psi value (ψ-value, measured in W/mK) quantifies the extra heat lost through a linear thermal junction, such as where a window frame meets a wall opening. Unlike U-values, which describe heat loss through flat building elements in W/m²K, psi values capture the additional heat flow caused by geometric discontinuities and material changes at edges and junctions. At window openings specifically, the psi value accounts for the fact that heat escapes faster around the perimeter of the frame than through the wall or the glazing alone. For a typical house with 60–80 m² of window perimeter (roughly 20–30 windows), even small errors in assumed psi values compound into measurable differences in whole-building heat loss — often 5–15% of total fabric losses.

What a Window Junction Psi Value Actually Measures

Also worth reading: How do you achieve compliant passive house thermal bridge detailing in modern construction? · What are the most effective thermal bridge mitigation strategies for urban building envelopes in extreme heat climates? · What is a psi value in thermal bridging and how does it affect SAP calculations?

When a window is installed in a wall, three heat flows interact: conduction through the wall, conduction through the frame and glass, and a concentrated 'funnelling' effect at the reveal where these elements meet. The psi value isolates that third component. It is calculated as the difference between the actual two-dimensional heat flow through the junction modelled numerically (typically with finite element software such as TRISCO, THERM, or Physibel) and the one-dimensional flows calculated from the U-values of the adjoining elements multiplied by their lengths.

The result is expressed in watts per metre of junction length per kelvin of temperature difference. A well-detailed insulated cavity wall with an extended cavity closer around a window might achieve ψ = 0.05 W/mK, while a poorly detailed steel lintel over the same opening could exceed 0.50 W/mK — a tenfold difference for what looks like a minor construction detail. Because the psi value multiplies by both the junction length and the temperature difference across it, a 6-metre perimeter around a large picture window with a bad detail can lose as much heat annually as several square metres of uninsulated wall.

It is worth being critical here: psi values are often treated as precise engineering constants when they are actually model outputs dependent on assumptions about insulation conductivity, frame position, and internal surface resistances. Two assessors modelling the same junction can produce values differing by 10–20% simply through different mesh densities or boundary conditions. Treat published values as indicative ranges, not gospel.

Typical Psi Values for Common Window Installation Details

The table below summarises representative accredited or conventionally accepted psi values used in UK SAP assessments and Passivhaus design. These are indicative; project-specific calculation always beats generic defaults.

External wall insulation delivers the best window junction performance because the insulation layer wraps continuously over the frame edge, effectively eliminating the reveal as a thermal weak point. Internal insulation performs worst because it pushes the structural wall into colder territory and concentrates heat loss at reveals, which is why retrofitting solid walls internally demands careful junction design to avoid condensation risk on cold reveal surfaces.

Why Window Junctions Matter More Than Most Designers Assume

Window openings are among the few junctions that appear repeatedly on every elevation of every building. A detached house might contain 25 linear metres of head detail, 25 metres of sill, and 40 metres of jambs once all openings are counted. Multiply those lengths by realistic psi values and the aggregate becomes substantial: using poor details averaging 0.20 W/mK across 90 metres gives 18 W/K of additional heat loss, whereas enhanced details averaging 0.04 W/mK give only 3.6 W/K. In a heating-dominated climate with roughly 2,500 degree-days equivalent temperature difference, that 14.4 W/K difference translates to around 36 kWh per year per kelvin — hundreds of kilowatt-hours annually, and proportionally more in colder regions like Canada or Scandinavia.

There is also a comfort and condensation dimension. The fRsi factor (temperature factor) at a junction describes how close the internal surface temperature stays to room temperature. Building standards typically require fRsi ≥ 0.75 for dwellings to limit mould risk behind furniture and curtains — exactly where window reveals sit. A junction can pass an energy calculation while still failing a condensation check, so both metrics must be assessed together. This dual requirement is why Part L compliance reports and building performance gap studies keep flagging junctions: the as-built detail frequently diverges from the modelled one, and the gap shows up first as cold, damp reveals rather than as a line item on an energy bill.

How Psi Values Are Calculated and Accredited

The standard method is two-dimensional numerical modelling following BRE IP 1/06 and BR 497 conventions in the UK, or ISO 10211 internationally. The modeller draws the junction geometry, assigns material conductivities, applies boundary temperatures (typically 20°C inside, 0°C outside for UK conventions), and solves the steady-state heat equation across a fine mesh. The psi value is then extracted from the heat flux results after subtracting the one-dimensional components.

Three routes exist for obtaining values. First, use Accredited Construction Details (ACDs), which carry default psi values accepted by SAP without further proof — though many were published years ago and reflect conservative assumptions. Second, commission bespoke calculations from a thermal modeller, costing roughly £150–£400 per junction in the UK market, with discounts for batches; a full set of 8–12 junctions for a housing typology typically runs £1,500–£3,000. Third, adopt values from certified detail catalogues such as the Passive House Institute's certified components database, which publishes rigorously verified junctions for specific wall build-ups and frame types. For Passivhaus certification, junction-specific psi values are mandatory, not optional, and the space heating demand target of 15 kWh/m²a cannot be met with default pessimistic values on most designs.

Practical Steps to Get Window Junctions Right on Your Project

Start at design stage, not on site. The single highest-leverage decision is frame position within the wall depth. Moving the frame outward to align with the insulation layer shortens the thermal pathway and can halve the psi value compared with a frame set deep in a cold reveal. For external insulation schemes, aim for the frame to sit flush with or slightly proud of the insulation face, with the insulation overlapping the frame edge by 30–50 mm where the system allows.

Second, specify thermally broken or insulated lintels at heads. Standard catnic-style steel lintels are notorious psi offenders; insulated variants cost £20–£80 more per unit and routinely cut head junction values from above 0.3 down below 0.1 W/mK. Third, close cavities properly with insulated closers rather than dpc-only closers at jambs. Fourth, insulate under sills and at reveals where the build-up permits. Fifth, ensure airtightness tapes and membranes are continuous from the frame to the air barrier line — airtightness failures at window perimeters are among the most common causes of the performance gap identified in post-completion testing, with blower-door tests regularly showing 20–40% worse results than design targets when sealing is left to site improvisation.

On site, the critical control points are: correct frame position before fixing, continuity of insulation at reveals (no gaps behind plasterboard returns), fully filled cavity closers, and taped membrane joints inspected before reveals are closed up. Photograph every junction before covering — retrospective verification is essentially impossible once plasterboard goes on.

Comparing Compliance Routes and Their Trade-offs

No route is universally superior. Defaults are free but pessimistic enough to sink a marginal Passivhaus design; bespoke modelling pays for itself on larger schemes where repeated across multiple plots; certified catalogues offer the best accuracy-to-effort ratio if your wall build-up happens to match a listed detail. Thermographic surveys are diagnostic rather than predictive — they confirm failures after the fact, which is useful for warranty claims but useless for prevention.

Common Mistakes That Inflate Window Junction Heat Loss

The most frequent error is assuming the frame's BFRC or declared U-value tells you anything about the junction. A 1.2 W/m²K frame installed badly can perform worse overall than a 1.4 W/m²K frame installed well, because installation quality dominates at the perimeter. Second, designers routinely forget that psi values feed directly into SAP and PHPP energy models — substituting better details mid-project invalidates earlier compliance calculations and requires re-submission.

Third, site teams substitute materials silently: swapping an insulated cavity closer for a plain plastic one, or omitting reveal insulation 'because it wouldn't be seen', destroys modelled performance invisibly. Fourth, airtightness and thermal bridging are treated as separate trades when they intersect physically at the same junction; a perfectly insulated reveal with a torn vapour-open tape still fails. Fifth, and most subtly, some assessors cherry-pick optimistic psi values from unrelated details to pass compliance — a practice that contributes directly to the documented building performance gap, where measured consumption exceeds predictions by 2–5 times in poorly governed projects. Regulators have responded: England's Part L uplifts from 2022 onward tightened fabric requirements and pushed assessed psi values toward verified figures, and further tightening expected through 2025–2026 continues this trajectory.

When to Act and What It Costs

Act during RIBA Stage 3–4 (detailed design), before construction drawings freeze. Retrofitting junction quality after tender means either accepting defaults or paying variation costs. The economics favour early action decisively: upgrading to insulated lintels and proper closers across a typical 3-bed house costs £300–£800 in materials and adds perhaps half a day of labour, while the resulting psi improvements can reduce modeled heat loss by 8–15%, cut boiler or heat pump sizing, and improve EPC bands. Commissioning bespoke calculations makes sense when the value of avoided plant downsizing plus avoided compliance failure exceeds the £1,500–£3,000 fee — generally true for developments of five or more units sharing a repeating detail set.

For retrofits, timing aligns with any window replacement or external insulation programme. Replacing windows without addressing reveals wastes much of the benefit: new glazing in old cold reveals leaves the perimeter as the dominant remaining weakness, and occupant complaints about draughts and condensation frequently follow. Pairing window replacement with reveal insulation and taped airtightness lines typically adds 10–20% to the window package cost but captures the full performance available.

The Bottom Line on Psi Values at Windows

Psi values at window junctions are small numbers with outsized consequences. They sit at the intersection of energy compliance, condensation risk, comfort, and buildability, and they are uniquely vulnerable to the gap between drawing and site reality. The defensible approach is straightforward: choose frame positions aligned with insulation layers, specify insulated lintels and closers explicitly in schedules (not just notes), verify junction-specific values for anything targeting better-than-building-regulation performance, inspect and photograph junctions before closing them up, and pressure-test the finished envelope against its target. None of this is exotic; nearly all of it is cheap at design stage and expensive to fix afterwards. Projects that treat junction detailing as a first-class design activity consistently outperform those that discover thermal bridging through a thermographer's camera in January.