Handbook of Radon.
54. Testing of building sites for radon levels.
There is currently no proven method of predicting indoor radon levels from pre-construction tests, and attempts to do this in other countries have produced poor correlations. Indeed, measurements of radon levels in the ground can be quite misleading. Radon is almost ubiquitous in rocks and soils in the UK, as elsewhere.
There are many factors involved, and a reading in the ground that appeared high can be really little indication of a potential problem. A more reliable guide might be the radon levels in many nearby buildings and combined with a knowledge of local geology. However, adjacent buildings can often have different indoor levels, by a factor of 100 or even 300.
At the moment, testing of new-build sites for radon potential does not appear to be useful in the UK. In other countries, classification systems have been developed for ground, based in part on in-situ radon measurements, but these are likely to find wide application only where the ground is fairly uniform over substantial areas, a situation not found in the UK so far as can be judged by the detailed distribution of high radon houses.
In Sweden, a system is used under which ground is classified as being of high, normal or low radon risk. Both permeability and radon in soil gas are considered. Similar schemes are operated in parts of the USA, but based more on broad geological considerations and results from measurements in existing houses than on tests of the ground. (see Section 9 also).
In both countries the requirements for new-build housing may be determined by this classification, with full radon resistant construction being reserved for the 'hottest' areas. A unique set of ground testing or building construction requirements is unlikely to develop in the USA especially, because of the local nature of Building Code enforcement.
It is known that high radon houses can be found at one end of a building site, yet only a few metres away houses of similar construction all have low levels. It seems inescapable that ground tests at one end of a site, at some arbitrary depth, could be useless for predicting what would happen even a few metres away. Also, tests at different depths on the same plot can show marked differences, and there must be doubt as to the most appropriate depth for testing given the variations sometimes necessary in house foundations as a consequence of conditions discovered only at time of construction.
It would be an attractive prospect to be able to predict radon levels in new buildings by inexpensive tests on the pre-construction site. The problems are not those of instrumentation, but of inherent variability of the ground. What may well be possible is to classify sites as low risk, thus avoiding the need to incorporate radon precautions, and based on assessment of homogeneity (deduced from maps) and uniformly low results from soil gas sampling as well as permeability.
KEY FACTS:
Testing of green-field sites for radon cannot be recommended in the UK as a method of predicting high indoor levels. This position may change, but some doubt may always remain in respect of sites where the ground conditions are inhomogeneous. Design and construction standards can also much influence indoor radon levels, as can occupant behaviour, especially in naturally ventilated buildings.
Radon levels in every building may never become predictable to great accuracy, (see Sections 55 and 56) but prediction of high and (especially) low risk sites should certainly be possible in some cases, and with possible savings in construction costs.