Radon, and the three months that start now

Radon is the one building risk that comes from underneath. It is a natural radioactive gas produced by uranium in rock and soil, it moves up through the ground, and it collects in the enclosed spaces above. Which means it is not really a property of the building at all. It is a property of the ground the building sits on, and of how tightly that building is sealed.

That second half is why the end of August matters. Radon concentrations vary through the year and are usually at their highest in winter, when doors stay shut, ventilation drops and the heating draws air up through the floor. A measurement has to run for three months to average out that variation. Detectors placed now sit through September, October and November. Detectors placed in January have already missed part of the season they are meant to capture.

The two levels, and which one applies

For homes, the UK Health Security Agency sets an Action Level of 200 becquerels per cubic metre and a Target Level of 100 Bq/m³. Both are annual averages. The measurement is done with two detectors, one in a bedroom and one in a living area, left in place for three months, with a seasonal correction applied to give the annual figure (UK Health Security Agency).

For workplaces the number is different and so is the law. The Ionising Radiations Regulations 2017 apply where radon exceeds an annual average of 300 Bq/m³, at which point the employer has to restrict exposures and the Health and Safety Executive has an enforcement interest (Health and Safety Executive).

The gap between 200 and 300 catches people out. A converted ground floor with residents above can sit under two regimes at once, and the figure that clears the workplace threshold does not clear the domestic one.

The duty starts before the threshold

The threshold is not where the obligation starts. The Management of Health and Safety at Work Regulations require an employer to assess health and safety risks, and the HSE is specific about what that means for radon: measure ground floor rooms of workplaces in radon Affected Areas, and measure all occupied below-ground workplaces wherever they are. A basement plant room, a records store, a staff area under a shop. Location on the map does not excuse those.

An Affected Area is defined by UKHSA as one where 1% or more of homes are at or above the Action Level. Cornwall, Devon, parts of Somerset, the Derbyshire Peak, Northamptonshire, parts of Wales and Grampian carry the highest potential, but the maps are indicative and the variation is local.

The HSE makes one point that is worth reading twice: results from neighbouring properties cannot be used in your risk assessment. Radon concentrations differ substantially between buildings that sit side by side on the same geology, because what actually decides the figure is the route in, the floor construction, the underfloor void, the cracks, the service entries. You cannot borrow a number from next door, and you cannot infer one from a map.

Why it is worth the three months

Radon is linked to over 1,100 deaths from lung cancer each year in the UK (UK Health Security Agency, citing Radon and Public Health, Advisory Group on Ionising Radiation, 2009). It is not a hazard that announces itself. There is no smell, no stain on the ceiling, no failed inspection. The only way anyone knows is because somebody put detectors in and waited.

The remedy, where a reading comes back high, is usually unglamorous and cheap by comparison. Better underfloor ventilation, sealing the obvious routes in, or a radon sump with a small fan. The expensive part is not the fix. It is discovering, years later, that nobody ever measured.

What the record has to show

Three things, in the order an inspector would ask for them.

That you considered it. A dated risk assessment that names radon and says why the building was or was not measured, referencing the radon potential of the location and whether any occupied space is below ground.

That the measurement was valid. Which detectors, from which validated laboratory, in which rooms, over which three months, and what seasonal correction was applied. A one week reading from a hardware shop device is a screening exercise, not a result.

That anything found was acted on. What the level was, what was done, and the confirmatory measurement afterwards showing the level came down.

Where SAMRISK fits

Radon sits in the awkward gap between the building and the land, which is exactly the gap SAMRISK was built to close.

The ground conditions and the radon potential of the site belong with the rest of the land and site record, alongside the drainage and the buried services. The assessment itself is a dated risk assessment with an owner, not a paragraph in a survey nobody has opened since. Where a measurement is due, or a retest after remedial work, the date sits on the compliance calendar rather than in somebody's diary. The laboratory certificates and the before and after readings live with the building's other documents, so the whole account can be produced at once. And where a high reading has been found and not yet fixed, it stays visible as a site risk until it is.

The measurement takes three months whenever you start it. Starting it this week is the difference between a result that covers the winter and a result that describes the spring.