A replacement rooflight or roof lantern in an existing dwelling in England must achieve a whole-unit U-value of 2.2 W/m²K or better under current Part L. New-build work is governed by the whole-dwelling SAP calculation rather than a single element figure, though 2.2 is treated as the backstop. A good thermally broken aluminium lantern with argon-filled double glazing and a warm-edge spacer achieves 1.2 to 1.4, comfortably inside the requirement.
So the regulation is not the hard part. The hard part is making sure the number you are quoted is the same kind of number as the one in the regulation, because two suppliers can both say “1.1” and mean entirely different things.
Whole unit against centre pane
This is the distinction that matters and it is routinely blurred in sales literature.
Centre-pane U-value measures heat loss through the middle of the glass only. It ignores the spacer bar at the edge and ignores the frame entirely. It is a real number and it flatters every product it is applied to.
Whole-unit U-value measures the complete lantern: glass, spacers, glazing bars, ridge, eaves beam, the lot, area-weighted. It is what Building Control assesses and it is always the worse of the two figures.
On a typical lantern the gap is substantial. A unit with a centre-pane value of 1.0 might have a whole-unit value of 1.4 once the frame is included, and on a small lantern with proportionally more frame it can be worse still. If a quote states a U-value without saying which one, assume centre pane and ask.
| Specification | Whole-unit U-value (W/m²K) | Notes |
|---|---|---|
| Lantern from the mid-2000s, no thermal break | 2.8 to 3.2 | What most failing units on Suffolk roofs achieve today |
| Part L replacement threshold, England | 2.2 | The legal floor, not a target |
| Double, argon, warm-edge, thermally broken aluminium | 1.2 to 1.5 | Our standard specification |
| Double, solar control, argon, warm-edge | 1.2 to 1.4 | Default on south and west facing lanterns |
| Triple, argon, warm-edge | 0.9 to 1.1 | Roughly 40% heavier glass, feeds into structure |
Typical values for aluminium lanterns of 2.0 x 1.0m and above. Smaller units carry proportionally more frame and perform slightly worse.
The vertical plane convention
Rooflight U-values are assessed in the vertical plane, which surprises people given that a lantern is anything but vertical. It is a testing convention rather than a physical claim, and it exists so that rooflights can be compared against each other on a consistent basis. A lantern installed at 25 degrees will not perform identically to its declared figure in the real world, which is one reason the regulation sets 2.2 rather than the 1.4 that vertical windows are held to.
The practical implication: do not try to compare a rooflight U-value directly against a window U-value and conclude the rooflight is poor. They are measured to different conventions.
What actually gets you a good number
- A polyamide thermal break in the frame. This is an insulating strip running through the aluminium profile separating the cold outer face from the warm inner face. Without it, aluminium conducts heat straight out and drips condensation onto the worktop in winter. Two lanterns can look identical in a photograph and differ by a full point of U-value on this alone.
- A warm-edge spacer. Old aluminium spacer bars conduct round the unit edge and create the cold line where condensation starts. The upgrade to warm edge is cheap and it is the best value item on the whole specification.
- Argon fill and a soft-coat low-e layer. Standard on any decent unit now, but worth confirming rather than assuming.
- An insulated upstand. This one is not on the lantern’s data sheet at all, and it matters. A cold-bridged upstand undoes a good lantern, because heat leaves round the perimeter regardless of what the glass does. Our page on roof lantern upstands covers it.
Ask for three figures on any quote. Whole-unit U-value, g-value, and light transmission. A supplier who can produce all three has thought about the specification. A supplier who offers only a centre-pane number has not, or would rather you did not look closely.
What this means locally
The practical driver in Suffolk and north Essex is the age profile of the existing stock. A great many lanterns went onto extensions between 2003 and 2010, and those units sit somewhere between 2.8 and 3.2 W/m²K. Replacing one with a modern unit at 1.3 more than halves the heat loss through that opening, which is a noticeable difference in a kitchen with a flat ceiling on a cold morning.
Retrofit lantern work is notifiable to Building Control in its own right, because it is both a structural alteration and a thermal element change. We handle the Building Control notification for you: retrofit work goes through a building notice or full plans application to Ipswich Borough, East Suffolk, Mid Suffolk, Babergh or Colchester Building Control. We tell you which route your job needs at survey, and the declared U-value forms part of what gets signed off.
Related questions
- How do you choose glazing for a roof lantern? puts the U-value in its decision order.
- When should a roof lantern be replaced? covers thermal performance as a trigger.
- Do you need planning permission for a roof lantern? explains where Building Regulations apply regardless.
See also condensation on roof lanterns and aluminium roof lantern installation, under roof lantern installation.
Get the real number
Every quote we issue states the whole-unit U-value for the specific unit you are buying, so you can compare quotes on the same basis. Request a fixed quote for your property in Ipswich, Suffolk or north Essex.
Thinking about this job? We survey across Ipswich, Suffolk and north Essex and give you a fixed price before any work starts, with a 10-year workmanship guarantee.