Roof lantern condensation and how to avoid it

Condensation is simple to explain and easy to underestimate. Air holds water as invisible vapour, and warm air holds far more of it than cold air. Cool a parcel of air down and at some point it can no longer keep all that vapour in suspension. The temperature where it gives up is the dew point. Any surface colder than the dew point of the air touching it will grow a film of water. That is all condensation is: warm damp air meeting something cold.

A roof lantern gathers three disadvantages in one place. It sits at the very top of the room, and warm moist air rises and stratifies, so the wettest, warmest layer in the house ends up pressed against the glass. The glass itself is glazing, which even at a good specification is colder than any plastered wall or ceiling around it. And on a clear night the sky is the coldest thing a roof can see, so the outer face of the lantern radiates heat straight up into space and the whole unit chills faster than a wall window ever would. Put those together and the lantern becomes the first surface in the room to drop below dew point and the last to climb back above it in the morning.

This is why lanterns and flat rooflights get a reputation for streaming when ordinary windows in the same house stay clear. It is not the product failing. It is geometry and position doing what they always do.

Before you can avoid it, you have to know which of three completely different things you are looking at, because they have three different causes.

On the inside of the inner pane

Water on the room side of the glass, the side you can wipe, is interior condensation. It is the common one and the one this page is about. It means warm indoor moisture has found a surface below its dew point. It is not a fault in the unit. It is a conversation between the glass temperature, the humidity in the room and the airflow, and all three of those are things you can change.

On the outside of the outer pane

Water on the outer face, usually a dew that appears overnight and burns off within an hour of daylight reaching it, is exterior condensation. Counterintuitively this is a sign of good glass, and it gets its own chapter below because so many people mistake it for a defect and worry needlessly.

Between the panes

Misting or water droplets sealed inside the cavity, where you cannot wipe them off either face, means the perimeter seal of the sealed unit has let go and the inert gas and desiccant have failed. That glass will never clear and only gets worse. That is the one case here that is genuinely a failed component, and the answer is a new sealed unit, which is replacement work rather than anything you can treat from a ladder. If your lantern is misting between the panes, the glazing has reached the end of its life and wants replacing.

You cannot lower the dew point of the air unless you know what is loading it with water. An average household puts a surprising volume of moisture into the air every day, and most of it is produced in exactly the rooms people put lanterns over.

• Cooking and boiling a kettle: a busy kitchen can release several litres of water vapour across a day, and a lantern over a knocked-through kitchen and dining space sits directly in the plume.
• Drying laundry indoors, especially on an airer under the brightest part of the room, which is often right beneath the lantern.
• Showers and bathing, if the wet room is anywhere near an open-plan space that shares air with the lantern.
• People and pets simply breathing. A family of four gives off well over a litre of water overnight.
• Unvented tumble dryers, fish tanks and a large number of houseplants, all of which quietly raise background humidity.

The pattern across south Essex housing makes this worse in one specific way. The classic 1930s semi, of which the area has an enormous stock, has usually had its back rooms knocked through into a single kitchen and dining space under a rear extension, and the extension roof is where the lantern goes. That puts the biggest single moisture source in the house, the cooking, directly under the coldest surface in the house, the glass, with the two connected by rising warm air. It is a well-designed condensation machine, and the fix is to break the chain at more than one point.

The single most effective thing you can do is keep the inner pane warm, because a warmer surface has to meet damper air before it wets. Glass temperature is governed by the U-value of the unit, the gas in the cavity, the low-emissivity coating and, at the edges, the spacer bar. The table below shows the approximate temperature of the centre of the inner pane on a cold night, taken at 20 degrees indoors and 0 outside, and the indoor relative humidity at which that surface starts to run. These are calculated figures for guidance, not a manufacturer’s warranty, but the shape of them is what matters.

Read across the bottom rows. A basic unit starts sweating once the room passes about 62 per cent humidity, which a kitchen reaches most evenings you cook. A good argon-filled double pushes that trigger up towards 82 per cent, which the same room might only touch while a pan is boiling. A triple unit holds out until nearly 88 per cent. You have not stopped the moisture, but you have raised the bar the room has to clear before the glass wets, and in practice that is the difference between a lantern that streams every cold morning and one that stays clear on all but the worst nights.

One detail the centre-pane figures hide is the edge. The coldest strip of any sealed unit is the perimeter, right against the spacer bar, and a plain aluminium spacer conducts heat out of that edge and drops the local temperature by a couple of degrees. That is why the first condensation you ever see on a lantern is a thin band around the edge of each pane, not a film across the middle. A warm-edge spacer, made from a low-conductivity composite rather than aluminium, lifts that edge temperature and is one of the cheapest specification upgrades that actually earns its keep on a lantern. Ask for it by name.

The same glass that keeps a winter lantern dry also does the summer job this business is built around. A south-facing lantern over an Essex garden is collecting a great deal of solar energy through the warm months, and the solar-control glazing that holds that heat out in July is a multi-coated low-E unit that runs warmer on the inner face in January as well. Specify for the aspect and you tend to solve both problems with one pane.

Warm glass raises the humidity the room can tolerate before it wets. Ventilation lowers the humidity in the first place. You need both, because glass alone cannot cope with a genuinely wet room and ventilation alone leaves the coldest surface exposed on the worst nights.

A roof lantern has one advantage no wall window can match: it opens at the very top of the room, exactly where the warm damp air has already gathered. An opening vent in a lantern clears stratified moist air far faster than a window at head height, because it is venting the wettest layer directly rather than stirring the room to reach it. On a high open-plan ceiling where nobody is going to reach a manual pole twice a day, an electric opening rooflight on a humidity or rain sensor will crack itself open when the air gets damp and shut when it rains, which is the difference between ventilation that is specified and ventilation that never actually gets used.

Around the glass, the ordinary habits still count. Run the cooker extractor while you cook and for a while afterwards, not just when the pan boils over. Keep trickle vents open through the winter rather than taping them shut against a draught. Give a bathroom fan the overrun it was designed for. Dry washing outside, or in a room with the door shut and its own extraction, rather than on an airer under the lantern. None of this is dramatic, but each one takes a slice of water out of the air that the glass then does not have to fight.

A lantern is not just glass. It sits on an upstand, a raised kerb built up from the flat roof deck, and it meets the ceiling through a frame and a plastered reveal. Every one of those junctions is a chance for a cold bridge, and a cold bridge is a condensation site whatever the glass does.

The upstand is the one people get wrong most often. If it is built as a bare timber or a thin uninsulated kerb, the inside face of it runs cold and you get condensation and, in time, damp staining on the plaster where the lantern meets the ceiling, even when the glass itself is bone dry. A properly detailed upstand is insulated on the outside and continuous with the roof insulation below it, so the warm plastered reveal on the inside never falls to dew point. This is a detail settled at installation and difficult to change afterwards, which is why it belongs in the survey conversation and not in a later panic about a wet corner.

The frame matters for the same reason. An aluminium lantern bar with no thermal break is a metal path from the cold outside to the warm inside, and it will collect condensation along its length like a cold water pipe. A thermally broken frame, where an insulating section separates the outer metal from the inner, keeps the inside face warm enough to stay dry. When we detail a roof lantern installation, the upstand insulation and the frame’s thermal break are specified alongside the glass, because a warm pane in a cold frame on a cold kerb still gives you a wet ceiling.

Every winter, someone looks up at a lantern on a still, clear morning and sees the outer glass fogged with dew, and assumes the unit has failed. It has not. External condensation forms precisely because the glass is good.

A well-insulated unit does its job by refusing to let heat pass from inside to out. On a clear night, the outer pane radiates its warmth up to a cold sky and, because the excellent glazing behind it is not feeding any heat forward to replace it, the outer surface drops below the dew point of the damp night air and collects dew, exactly as a car windscreen or a lawn does. A poor, heat-leaking unit stays warmer on the outside and does not do this, which is a roundabout way of saying the fog on the outside of your lantern is the glass proving it is keeping the heat in. It clears within the hour once daylight reaches it. This is more common near the estuary, where clear nights and humid coastal air over open water give strong radiative cooling, and it is most visible on precisely the high-performance units you would want to buy. It needs no action at all.

You can prevent almost all interior condensation at the point of specification, before anyone orders glass. Here is what to settle with whoever is quoting.

1. “What is the whole-unit U-value?” Lower keeps the inner pane warmer and raises the humidity the room can tolerate. Ask for the whole-unit figure, not the flattering centre-pane one.
2. “Is there a warm-edge spacer?” The perimeter is where condensation starts. A composite warm-edge spacer instead of aluminium lifts that edge temperature for very little cost.
3. “Is the frame thermally broken?” A metal bar with no insulating break will wet along its length whatever the glass does.
4. “How is the upstand insulated?” The kerb should be insulated continuously with the roof, so the internal reveal never runs cold. This is fixed at installation.
5. “Is there an opening vent, and can it be automatic?” A vent at the top of the lantern clears the wettest air where it collects. On a high ceiling put it on a sensor so it actually gets used.
6. “Does the glass suit the way this roof faces?” On a south-facing plane the solar-control glass that keeps summer heat out is also a low-E unit that runs warmer inside in winter. One specification, both problems.

Do those six things and condensation stops being something that happens to your lantern and becomes something you designed out of it. If you want the glass, upstand and ventilation worked out for your own roof and the way it faces before you commit, ask us for a specification and we will set the numbers against your room rather than against a catalogue.