Whole unit U-value against centre pane

Centre pane

Measures heat loss through the glass at the centre of the pane, away from the edges. It considers the panes, the cavity, the gas fill and the coatings.

It ignores the spacer bar and the frame entirely, which are exactly where heat escapes. It is always the better number, and it is what marketing tends to quote.

Whole unit

Includes the glass, the spacer and the frame, weighted by area. It is what the building actually experiences and what Building Regulations use.

Always worse than centre pane, frequently by 0.4 or more, and on a small unit with a lot of frame relative to glass the gap is wider still.

Why the gap

Because the edges are the weak part. The spacer bar conducts around the perimeter of every pane, and the frame is a solid element of much poorer performance than the glazed area.

On a small rooflight the perimeter is large relative to the area, so the frame and spacer dominate. On a large one the glass carries more of the average and the two figures converge.

Comparing quotations

Ask which is quoted. A rooflight advertised at 1.0 and another at 1.4 may well be the same performance described two ways.

If a quotation will not state whole unit, that tells you something.

Typical figures

• Modern double, low emissivity and argon: centre pane around 1.0, whole unit 1.2 to 1.4
• Triple glazed: centre pane around 0.6, whole unit around 0.8

The number nobody quotes

The kerb, which is not part of the unit at all and frequently loses more than the glass. There is no figure for it on any quotation.

Two numbers, one product, and only one of them is regulated

This is the single most common source of confusion in rooflight quotations, and it is worth understanding precisely because it makes two quotations look further apart than they are, or closer together than they are, depending which way round the figures have been given.

Centre pane is a measurement of the glazing alone, taken away from the edges of the unit. It excludes the spacer bar and it excludes the frame entirely.

Whole unit is a measurement of the assembled rooflight: glass, spacer and frame, weighted by the proportion of the total area each occupies.

Whole unit is always the worse number. Frequently by 0.4 or more, and on a small rooflight the gap is wider still.

Why the gap exists

Because the middle of a sealed unit is the best insulated part of a rooflight and the perimeter is the worst.

In the middle, heat has to conduct through glass, then cross a gas filled cavity where a low emissivity coating reflects most of the infrared back into the room, then conduct through glass again. That is a genuinely good insulator for its thickness, and it is what centre pane measures.

At the perimeter, the two panes are connected directly by the spacer bar. A traditional aluminium spacer conducts extremely well, bridging warm pane to cold pane around the entire edge. Around that sits the frame, and the frame is a solid element with its own conductivity.

So centre pane describes the part of the rooflight that performs well and excludes precisely the part where the heat gets out.

Why it matters more on a rooflight than a window

Two reasons, and both make the gap larger.

Rooflights are small. A 780 by 1400 roof window has a lot of perimeter relative to its glazed area. The smaller the unit, the greater the proportion of frame and spacer, and the wider the gap between centre pane and whole unit.

On a large picture window the two figures converge. On a modest rooflight they diverge.

Rooflights face the sky. They radiate toward the coldest thing available, particularly on a clear night, so their edges get colder than a wall window’s do. That is why frost forms on rooflights before it forms on windows, and it is why edge effects show up more visibly.

What the regulations use

Whole unit. Part L expects a whole unit U-value no worse than 2.2 for a replacement rooflight in an existing dwelling.

Any current double glazed unit with a low emissivity coating, argon fill and a warm edge spacer lands at 1.2 to 1.4 whole unit, so the threshold is cleared comfortably. The regulated question is rarely the difficult one.

The difficult one is comparison, and that is where the two figures cause trouble.

The comparison problem, worked through

Supplier A quotes a rooflight at 1.0. Supplier B quotes what appears to be a similar unit at 1.4. On the face of it A is substantially better.

If A has quoted centre pane and B has quoted whole unit, they may be the same product. A’s whole unit figure might be 1.4 or worse.

Neither supplier has necessarily done anything wrong: centre pane is a real measurement and it is commonly published. But comparing one against the other tells you nothing, and it will lead you to the wrong choice.

The rule is simple. Ask both for whole unit figures, in writing, on the quotation, alongside the model and size code. If one cannot or will not provide it, that is informative in itself.

Where the perimeter losses actually are

The spacer bar. An aluminium spacer creates a thermal bridge around the whole edge of the unit, and the glass edge runs several degrees below the centre. The visible symptom is condensation forming as a band around the edge of the pane on cold mornings while the middle stays clear.

That is extremely common on units fitted before the mid 2000s and it is regularly mistaken for a failed sealed unit, which it is not. A warm edge spacer in stainless steel, structural foam or a composite runs several degrees warmer and largely removes it, improving the whole unit figure by around 0.1.

The frame. Timber is a reasonable insulator, which is why quality roof windows use laminated timber cores with aluminium cladding outside: the cladding takes the weather and the timber does the insulating.

Aluminium framed products, including every lantern, need a thermal break: a section of low conductivity material interrupting the path from inside to outside. Without one the frame is a set of cold bridges holding glass, and it shows as condensation on the internal frame members long before the glass does anything.

The number that appears on no specification at all

Neither figure includes the installation, and on many rooflights the installation loses more than the unit does.

On a flat roof, the kerb. A 1200 by 1800 rooflight has six metres of perimeter, and an uninsulated timber kerb 150mm high presents very nearly a square metre of additional envelope, punching straight through the roof insulation, at a U-value several times worse than the glass above it.

In other words, a poorly detailed kerb can lose comparable heat to the entire rooflight sitting on it, and no glazing specification compensates.

On a pitched roof, the insulation collar. Without it the reveal runs cold enough for room air to condense on it, which is a heat loss and the commonest cause of water appearing on plaster.

And on both, air leakage. A vapour control layer that is not continuous and sealed to the frame loses more heat than conduction through a small area of glass does, and it carries moisture into the construction as well.

What to ask for, and in what order

The whole unit U-value, stated as whole unit, in writing on the quotation.

Confirmation of a warm edge spacer, which should be standard rather than an upgrade.

The g value and light transmission where the unit faces south or west, because on a sunward facing rooflight solar gain matters more to how the room feels than the U-value does.

And, separately from anything about the glazing, confirmation that the kerb is insulated on its outer face or that an insulation collar is included.

Those four lines tell you more about how the installation will actually perform than any comparison of headline numbers between brands.

A sense of proportion

None of this is an argument that U-values do not matter. It is an argument that they are frequently quoted in a form that cannot be compared, and that the differences people spend money chasing between one good unit and another are smaller than the differences created by details that appear on no data sheet.

A unit at 1.4 whole unit, fitted with an insulated kerb and a sealed vapour control layer, will outperform a unit at 1.2 fitted without them, in the room, every winter.

What each number describes

Centre pane describes heat loss through the middle of the glazed area, away from the edges. It is a property of the glass build up alone: the panes, the coatings, the cavity and the gas fill.

Whole unit describes the complete assembly as installed: the glass, the spacer, the frame, and the edge effects where they meet.

Both are legitimate figures. Only one of them describes what your roof actually loses.

Why they diverge most on a rooflight

The ratio of edge to area.

A large window has a lot of middle and comparatively little edge, so the centre pane figure dominates and the two numbers are close.

A rooflight is frequently small. A small unit has a great deal of perimeter relative to its glass area, so the spacer and frame carry a large share of the total loss, and the whole unit figure is substantially worse.

The consequence is important: the whole unit U value changes with size. A manufacturer quoting one figure for a range is quoting it at a particular size, and your size may be worse.

Ask for the figure at the size you are buying.

How this misleads in practice

Two units, both quoting an impressive centre pane figure. One has a warm edge spacer and a thermally broken frame; the other has an aluminium spacer and a simple frame.

On centre pane they look identical. On whole unit at a small size they are not close, and the difference is what you will feel in the room and see as condensation at the glass edge.

What still is not included

Even the whole unit figure stops at the frame.

The junction between the frame and the roof insulation is outside it entirely. An insulation collar, an insulated kerb, insulation carried continuously around the reveal, and a vapour control layer sealed to the frame are worth as much as the glazing specification in the actual building, and none of them appear in any published number.

A well specified unit in a badly detailed opening performs worse than an ordinary unit in a well detailed one.

The order to spend in

On a small rooflight: the perimeter detailing first, then a warm edge spacer, then a good coating and gas fill, then triple glazing if the unit is large enough for it to matter.

Reversing that order is common and it buys the least improvement per pound.