Draught-proofing works on air exchange rather than on materials

Two separate routes out

A heated building loses energy in two distinguishable ways. Heat conducts through walls, roof and floor, which is what insulation addresses. And warm air physically leaves the building and is replaced by cold air, which insulation does nothing about at all.

That second route is called infiltration when it is unintentional and ventilation when it is deliberate, and the physical process is identical. Every cubic metre of air that leaves at room temperature and is replaced at outdoor temperature costs the energy required to warm that replacement, and in a leaky building that happens many times an hour.

The distinction matters because the two problems have different solutions and different costs. Insulation is a materials job and it is priced by area. Draught-proofing is a detailing job, priced by attention, and it is usually the cheapest measure in the house per unit of energy saved.

What drives the air through

Air does not leave a building at random. Two forces push it, and understanding them tells you where to look.

The first is the stack effect. Warm air is less dense than cold, so a heated building sits at slightly higher pressure at the top and slightly lower at the bottom relative to the outside. Air is therefore pushed out through gaps high up and drawn in through gaps low down, and the effect strengthens as the temperature difference grows — which is to say, exactly when heating is most expensive.

The second is wind. Pressure on the windward face of a building is positive and on the leeward face negative, so wind drives air in one side and out the other. This is why a house feels draughtier on a windy day at the same outdoor temperature, and why exposed sites lose more than sheltered ones from an identical building.

Where the gaps actually are

People look for draughts at windows and doors, because that is where you can feel them on your face. The larger leakage paths are usually elsewhere and are not felt at all, because they are hidden inside the construction.

The floor perimeter, where floorboards meet the wall, is a classic route in older houses with suspended timber floors over a ventilated void. So is the gap around every pipe and cable passing through a wall or floor. Loft hatches are frequently the single largest hole in the top of a building, sitting exactly where the stack effect is pushing hardest. An open chimney is a permanent flue drawing heated air out of the room continuously.

Recessed light fittings, extractor fans without effective backdraught closures, and the void behind boxed-in services all connect conditioned space to unconditioned space. Sealing these is cheap and unglamorous and it is where the energy is.

The line between draught-proofing and over-sealing

This is the point at which honest advice has to include a warning. A building needs air exchange, for the occupants and for moisture. Cooking, washing, drying clothes and breathing all add water vapour to indoor air, and if that vapour is not removed it condenses on the coldest surfaces available.

The result is condensation, mould growth and, in some constructions, moisture accumulating within the fabric itself. Sealing a house tightly without providing controlled ventilation converts an energy problem into a building-health problem, and the second is considerably more expensive to fix.

The principle that resolves it is to build tight and ventilate right: eliminate uncontrolled leakage, and provide deliberate ventilation you can regulate, whether that is trickle vents, extract fans in wet rooms, or a mechanical system with heat recovery in a more thorough retrofit. What you are removing is randomness, not air. Anyone advising you to seal everything without mentioning ventilation is giving incomplete advice.

How far to take it

The first tranche of draught-proofing — loft hatch, floor perimeter, service penetrations, an unused chimney, letterbox and keyhole covers, a properly fitted door seal — is inexpensive, achievable by a competent amateur over a few weekends, and recovers its cost quickly in any climate where heating is used seriously.

Beyond that, the returns fall off and the work gets harder. Achieving genuinely low air leakage in an existing building requires a systematic approach, membranes and tapes at junctions, and ideally a pressure test to find where the remaining leakage is. That is a substantial retrofit rather than a weekend, and whether it pays depends on the climate and on what the building already achieves.

The measure has a natural floor as well. Once air exchange has been reduced to what the occupants and the moisture load actually require, there is nothing further to save, because the remaining ventilation is not waste. It is the point of the system.