Roof Lantern Installation

Condensation on a lantern, and what actually causes it

Water on a roof lantern gets reported as a leak more often than as anything else, and a fair proportion of the time nothing is leaking. There are four distinct places water appears on a lantern, they have four different causes, and only two of them are faults.

This page is about telling them apart from the ground, before anybody goes on the roof.

The four places water appears

Where the water is, and what it means
Where Cause A fault? Weather it happens in
Inside face of the glass Room humidity against a cold pane Usually not, unless persistent Cold, and worse overnight
Inside face of the frame and reveal Thermal bridge at the kerb Yes Cold and clear, not raining
Between the panes Failed edge seal Yes, and terminal Any, and it never clears
Outside face of the glass Dew on a well insulated unit No. The opposite Clear still nights, spring and autumn

Everything below is an expansion of that table.

Inside the glass: the ordinary case

Warm room air holds water vapour. The inside face of any glazing is the coldest large surface in the room. Where the air touching that surface cools below its dew point, water comes out of it onto the glass.

A little of this, in cold weather, in a kitchen or a bathroom, is normal physics rather than a defect. It appears overnight when the heating is off and the room has been closed up, and it clears through the morning as the room warms and the air moves.

When it stops being normal

It matters when it is persistent rather than overnight, when it runs down and pools on the frame rather than sitting as a mist, or when it appears on mild days as well as cold ones. Any of those points at humidity that is genuinely too high for the room rather than at the glass.

What changes it

Better glass helps, because a lower U-value means a warmer inner pane. Ventilation helps more, because it removes the vapour rather than raising the surface it lands on. In practice the two together are the answer and neither alone quite is.

On the frame: the one that is a fault

Water on the frame and running down the reveal, in cold clear weather, is the signature of a cold bridge at the kerb. The insulation stops at the bottom of the kerb, the kerb and the frame sit at close to outside temperature, and room air condenses on them.

Two things separate this from ordinary glass misting. It is on the frame rather than the glass, and it happens on cold nights rather than rainy ones. If your water follows the temperature and not the rain, and it is on the metal, this is what you have.

The mechanism, the diagnosis and the fix are set out in full on the thermal bridging page. It is a construction detail rather than a product fault, and it is not solved by changing the lantern.

The eaves beam of a roof lantern where it sits down on its kerb, seen close from inside, the thermal break and the glazing gasket running along the frame
Between the panes: the terminal one

Between the panes: the terminal one

A sealed unit is two panes with a sealed cavity between them, filled with dry gas. The seal at the edge keeps moisture out and gas in.

When that seal fails, moist air enters the cavity. It condenses on the inside faces of the panes in cold weather and clears in warm, and over time it leaves a permanent haze and mineral staining that never clears at all.

Nothing you do to the room affects it

That is the diagnostic. Ventilation, heating and dehumidifying all change condensation on the room side of the glass. None of them touch misting inside the cavity, because the cavity is not connected to the room.

It cannot be repaired

Processes exist that drill and dry a failed unit, and they address the appearance rather than the cause. The gas fill is gone, the thermal performance is gone with it, and the seal is still failed. The answer is a new unit, and on a lantern that is frequently a pane replacement rather than a whole unit if the frame is sound and reglazable.

Outside the glass: the one that means it is working

This is the most commonly reported and the most commonly misunderstood.

On clear still nights, mostly in spring and autumn, dew forms on the outside of a well insulated rooflight. The outer pane radiates heat to a clear sky and drops below the dew point of the outside air, and the moisture in that air condenses on it, exactly as it does on a car windscreen or on a lawn.

Why it happens on good units and not bad ones

A poorly insulated unit leaks enough heat from the building to keep its outer pane above the dew point. A well insulated one does not, because that is what insulation means. External dew on a rooflight is therefore evidence that the unit is doing its job.

What it looks like

The glass reads misty from inside first thing, and it clears within an hour or two of the sun getting onto it, or as soon as any breeze picks up. It is on the outside surface, so a finger on the inside of the glass finds it dry.

Nothing to be done, and nothing wrong

There is no fault to fix. It is more noticeable on a lantern than on a vertical window because a lantern faces the sky directly, which is where the heat is radiating to.

The finger test

Where you cannot tell which surface the water is on, and from a floor below a lantern that is common, the sequence is simple.

  1. If you can reach it safely, touch the inside of the glass. Dry glass with visible mist means the water is outside or between the panes.
  2. Note whether it clears. External dew clears within a couple of hours. Cavity misting comes and goes with temperature over days. Room side condensation clears as the room warms.
  3. Note where on the unit it sits. All over the glass points at the glass. Only at the perimeter and on the frame points at the kerb.
  4. Note the weather. Rain points at a leak. Cold and clear points at condensation of one kind or another.

Four observations, none of which needs a ladder, and between them they identify which of the four cases you have.

Close detail of a timber upstand kerb built around a rooflight opening on a flat roof, single ply membrane dressed up the face of the kerb and terminated under the frame
Where the vapour comes from

Where the vapour comes from

Every case except external dew is driven by how much water is in the air inside the house, and the sources are more concentrated than people expect.

  • Cooking, especially boiling and steaming without an extract hood venting outside.
  • Showering and bathing, which puts a large quantity into the air very quickly.
  • Drying laundry indoors. A single load released into a room is a substantial amount of water.
  • An unvented tumble dryer, which puts the entire moisture content of the load into the room.
  • People and, in some houses, a large number of plants or an aquarium.

A kitchen extension with a lantern frequently combines several of these in one sealed, well insulated room, which is why lanterns get blamed for a house-wide humidity level.

Ventilation, in the order that works

Removing vapour at source beats removing it from the room, and removing it from the room beats warming the surfaces it lands on.

Extract at source

A cooker hood ducted to outside, and an extract fan in a bathroom, remove vapour where it is generated. A recirculating hood filters grease and returns the water to the room, which for this purpose does nothing at all.

Background ventilation

Trickle vents, or an opening vent in the lantern itself, provide the slow continuous air change that a sealed modern extension does not get by accident.

Purge ventilation

An opening rooflight at the ridge of the lantern clears a room quickly, because it is the highest point and warm moist air collects there. It is the single most effective vent position in the room.

Heating patterns matter more than heating levels

Condensation is a surface temperature problem, and surfaces respond slowly.

A room heated hard for two hours in the evening and left cold overnight produces a large amount of vapour in the warm period and then presents cold surfaces to it. A room held at a lower steady temperature keeps the surfaces above the dew point continuously and frequently has less condensation on less fuel.

This is not advice about how to run your house. It is the reason two identical extensions with identical lanterns behave differently, which is otherwise mysterious to the person in the one that mists.

The Sussex coast makes it worse in one specific way

The air arriving here off the Channel is humid, and it is humid all year rather than only in winter.

That raises the outside dew point, which is why external dew on a rooflight is more common on this coast than it is inland. It also means that opening a window on a mild damp day brings in air that is already close to saturation, so ventilation does less than it would in a drier climate.

The practical consequence is that mechanical extract at source is worth more here than open windows, and that a lantern facing a clear Channel sky will show external dew on more mornings a year than the same unit in Horsham.

The inside of a trussed rafter roof space, W shaped webs and punched metal plate joints clearly visible at the intersections, daylight from a gable vent
What is not the cause

What is not the cause

Three things get blamed regularly and are almost never responsible.

The glass specification is rarely the problem, because any current unit clears the limiting standard comfortably and the difference between a good unit and an excellent one is small compared with the difference ventilation makes. The size of the lantern is rarely the problem either, because a larger cold surface collects more water but does not create the vapour. And the installer is rarely the problem when the water is on the glass rather than the frame, because the glass surface temperature is a property of the unit and the room.

Where the water is on the frame and the reveal, the installation detail is squarely the problem, and that is the one case where it is.

When to call somebody

Most condensation on a lantern needs a change to ventilation rather than a visit. It is worth calling when any of the following is true.

  1. Misting has appeared between the panes, which is a failed unit and will not recover.
  2. Water runs off the frame and reveal every cold night, which is a kerb detail rather than a habit.
  3. Staining or mould has started on the plaster at the reveal, which means it has been going on long enough to be doing damage.
  4. Water appears in rain from one wind direction, which is not condensation at all.

What we ask first on a lantern

We ask which surface the water is on, what the weather was doing, and whether it clears, before we quote anything, because three of the four causes are not fixed by replacing the unit. Where it is a failed seal we replace the glass or the unit. Where it is the kerb we open up and correct the detail. Where it is the room we say so plainly rather than selling you a lantern you do not need.

Rain leaks follow rain. Condensation follows cold. That one sentence settles most of it.

Thinking about this job? We survey across Worthing, West Sussex and the South Downs and give you a fixed price before any work starts, with a 10-year workmanship guarantee.

Get a fixed quote

Get a fixed quote

Tell us about the roof. We will tell you the price.

A survey costs you nothing and there is no obligation attached to it. We look at the roof build-up, the exposure and the opening, and give you one fixed figure, not an estimate that moves once work starts.

If your rooflight is leaking now, say so and we will prioritise the survey.

  • Fixed price, confirmed in writing
  • 10-year workmanship guarantee
  • Installed to current Building Regulations
  • Your roof is never left open overnight
Call Get a fixed quote