House dust

Why does a room nobody enters still need dusting every few weeks?

Shut the door on the spare room in March. Open it in April and run a finger along the sill: grey. Nobody has been in. Nothing has been moved. Nothing has been used.

That single observation disposes of the explanation most people have heard, which is that house dust is mostly dead skin. If it were, an unoccupied room would be the one place in the house that stayed clean. So two separate questions have to be answered, and the popular account answers neither: where did the material come from, and — the question people rarely think to ask — why is it still there?

Take the second question first, because it turns out to be the more interesting one.

A particle in still air falls, and it falls at the speed where drag balances weight. For small particles that speed is proportional to the square of the diameter, which makes the size range in a room enormously important. Put numbers on it. A particle ten micrometres across, at roughly the density of water, settles at about three millimetres per second — so it crosses a two-and-a-half-metre room in around fourteen minutes. Shrink it to one micrometre and the speed drops by a hundred: now it takes about a day. At a tenth of a micrometre, gravity has effectively given up; such particles wander by Brownian motion, stick to whatever they happen to touch, and mostly leave with the air.

So the room performs a sorting operation. Coarse particles are delivered to horizontal surfaces within minutes of arriving. Fine ones stay airborne long enough to be ventilated away. What we call dust — the visible grey film — is precisely the fraction that gravity hands over quickly.

And then it stays. This is where an unused room is not protected but exposed. A settled particle is held by adhesion, and for small particles that grip is large compared with their weight, so ordinary air movement will not lift them again. Footsteps, opening doors, sitting on a bed and shaking a duvet all resuspend dust and give it another chance to leave through the ventilation. A closed room offers none of that. The deposit accumulates undisturbed, because the only removal mechanism a room has for settled dust is somebody wiping it off.

Now the first question. If nobody is shedding, what is arriving? Two streams. A house is not airtight: wind pressure and the stack effect drive outside air through gaps in the envelope, and the exchange rate for ordinary housing is commonly quoted at somewhere between a fifth and one whole air change per hour — a recalled range, and one that varies enormously with construction and weather. That air brings soil minerals, pollen, soot, road wear and sea salt. Meanwhile the room's own furnishings shed continuously with no help from anyone: carpet, curtains, upholstery, bedding, cardboard and paper all release fibres simply by existing. And a closed internal door is not a seal, so air from the occupied part of the house drifts in carrying its cooking particles, its fibres and, yes, some skin.

Three things to correct, including one in my own account.

The skin claim first. Figures like "eighty per cent dead skin" circulate widely and are not supported by composition studies, which typically find fibres and outdoor mineral matter dominant, with skin a real but modest share. More importantly, the composition is not one number: it depends on flooring, pets, smoking, proximity to a road, and how the house is ventilated. Anyone quoting a single percentage for all houses is quoting something that does not exist.

Second, this is an accumulation problem, and it is worth being precise about how it differs from the household one. Clutter is also inflow beating outflow, but there the outflow exists — charity shops, bins, giving things away — and simply goes unused. Dust has no outflow. Nothing in the physics of a closed room removes a settled particle. Cleaning is not accelerating the outflow; cleaning is the outflow, the only one there is.

Third, my own simplification. The settling arithmetic above treats particles as smooth spheres of water-like density. Real house dust is dominated by fibres, flakes and irregular mineral grains, which have more drag for their mass and therefore fall more slowly than the calculation suggests, and denser mineral grains fall faster. So read those numbers as showing the enormous spread across sizes, not as predictions for any particular speck.

How you could refute this. Put two clean glass slides in the closed room: one flat on a table, one taped vertically to the wall beside it. Leave them a month, then compare under a lens. My account says gravitational settling dominates, so the flat slide should collect dramatically more. If both collected about equally, deposition is being driven by something else — air currents, electrostatics — and the story above is wrong. A second test separates the two sources: run a filtered air purifier in the sealed room. If deposition falls sharply, the dust is arriving through the air; if it barely changes, the furnishings are shedding straight onto the surfaces beside them.

• Why the particle sizes that settle least readily are also the ones that reach deepest into the lungs.
• How air-change rate trades off against filtration, given that ventilation both delivers dust and removes it.

Every term the collection defines is gathered in the glossary.