What is a warm-edge spacer?

The part itself, and where to find it in a unit you already own

Look at the edge of the glass in any double or triple glazed rooflight and there is a bar running round the perimeter, set back a little behind the frame, usually 12 to 20mm wide. That is the spacer. It sets the cavity width, it holds the desiccant that mops up the small amount of moisture sealed in at manufacture, and it carries the two seals: a butyl primary seal that keeps vapour out and gas in, and a structural secondary seal that holds the sandwich together.

On an older unit the bar is bright, mill-finished aluminium and it looks like metal because it is. On a modern warm-edge unit it is usually matt black or grey, sometimes a very thin folded stainless steel section, sometimes a solid plastic or foam profile. Colour is a hint rather than proof, because black anodised aluminium exists, but a bright silver bar in a rooflight is almost always the conventional type.

Why the material matters more than the width

The cavity between the panes is a good insulator. The spacer is a bridge straight across it. Aluminium conducts heat at roughly 160 watts per metre kelvin. Stainless steel is around 17. The thermoplastics and structural foams used for warm-edge work sit under 0.5, and in a folded steel or hybrid spacer the metal is reduced to a foil a fraction of a millimetre thick specifically to cut that path down.

What that does at the glass is shift the surface temperature of the edge band. On a cold night the inner face of the pane at the edge of a conventional unit can sit four to six degrees below the centre of the pane. Change the spacer and a good part of that gap closes, typically by two to four degrees at the coldest strip. Dew point in a normally occupied house sits close enough to those temperatures that two degrees is frequently the difference between a wet band on the glass every frosty morning and a dry one.

What the term means as a defined property

Warm edge is not a marketing word with no content behind it. The European convention sets a threshold based on the material properties of the spacer section, so a product qualifies only where the sum of the wall thickness multiplied by the conductivity of each material in the section falls at or below 0.007 watts per kelvin. Below that line, the calculation method also allows a better default psi value to be used in the whole-unit figure. Above it, the spacer is treated as conventional. Manufacturers state the classification in their technical data, and a supplier who cannot say which side of the threshold a product sits on has answered the question anyway.

What it is worth in a whole-unit U-value

The edge appears in the calculation as a linear term: the psi value, in watts per metre kelvin, multiplied by the total length of the glass perimeter. A conventional aluminium spacer in a timber or composite frame carries a psi around 0.08. A good warm-edge spacer in the same frame is nearer 0.03 to 0.05. Multiply that difference by the perimeter and divide by the area, and a typical roof window improves by somewhere in the order of 0.1 W/m2K on the whole-unit figure. That is a small number until you notice it costs almost nothing and applies for the whole life of the unit. The way the terms combine is on U-values and glazing specification.

Small units, lanterns and multiplied perimeter

Perimeter divided by area is the figure that decides how much the spacer matters, and it gets worse as units get smaller. A narrow 550mm roof window between two trusses on a Quedgeley estate has a great deal of edge for very little glass, so the linear term carries far more weight than it does in a large flat rooflight over a kitchen.

A roof lantern is the extreme case, because every glazing bar is two more glass edges. A lantern made up of ten panes has several times the perimeter of a single unit covering the same opening, so the spacer specification moves from a detail to one of the main decisions. Sightlines and bar counts are on roof lantern sightlines and glazing bar thickness.

Overhead glass, where the edge zone works harder

Two things make the roof harder than a wall. Warm humid air rises and collects against the ceiling, so the air actually touching a rooflight is warmer and wetter than the air touching a window downstairs, which raises the dew point at the surface. And on a shallow unit, water condensing at the edge band has nowhere to run, so it sits in the drainage rebate rather than running down a pane and away. A colder edge on a rooflight therefore produces visible standing water sooner than the same edge in a vertical window.

The lines that get the right spacer onto an order

Two entries do it. First, state the spacer by name and classification: warm-edge, with the product named, and the psi value declared. Second, ask for the whole-unit U-value calculated to the actual unit size with that spacer included, rather than a centre-pane figure. A supplier quoting only the glass number is quoting the part of the unit that has no spacer in it. Severn records the spacer type at survey and puts it on the order alongside the coating position, the gas fill and the laminate build-up, and the reasons for each are on the energy efficient rooflights hub.

Where it shows up across Gloucestershire

The units that report a cold spacer loudest in this county are the small ones in cold rooms. Conservation rooflights on Cotswold stone slate through Painswick, Bibury and Northleach have slim steel frames and modest glass areas, so the perimeter fraction is high and the frame is already a conduction path. Shower rooms and kitchens lit from above in Gloucester’s Victorian rear additions put the wettest air in the house against the coldest glass. And on the low ground through Tewkesbury and the vale, where river mist and clear nights arrive together, the edge band is often the only part of the pane that wets at all. Warm-edge does not stop condensation on its own, ventilation does that, but it moves the threshold in the right direction on every one of those roofs.