Representativeness of nationwide indoor radon measurement data base

Category Other
Group GSI.IR
Location International Geological Congress,oslo 2008
Author Valmari, Tuomas; Mنkelنinen, Ilona; Arvela, Hannu
Holding Date 08 October 2008

Radiation and Nuclear Safety Authority (STUK) has so far carried out over 100 000 indoor air radon measurements in Finnish dwellings using alpha-track detectors. The measurement results, together with the building type information forms filled in by the residents, are used for producing radon maps. They can also be utilised for studying the effects of soil type, foundation type and other building characteristics, radon preventive actions, weather conditions etc. on the indoor radon concentration. However, the data base is not a representative sample of all Finnish buildings because people suspecting a high radon level are more likely to carry out the measurement. Measurement activity is affected by location, as well as by foundation type and other building characteristics.
The representativeness of a sub-set of 38 000 measurements carried out before 1997 was studied by calculating average radon concentrations in 1 km x 1 km map grid and comparing them with building statistics. The averages of all the measurements were 259 Bq/m3 (detached houses), 265 Bq/m3 (flats) and 260 Bq/m3 (total). Half of the dwellings locate in those 1 x 1 km squares where the average radon is above 114 Bq/m3. These higher-radon squares contributed to 68% of the number of measurements. The building density-weighed radon averages were 162 Bq/m3 (detached houses), 114 Bq/m3 (flats) and 158 Bq/m3 (total). These values are only moderately higher than the respective values of 145 Bq/m3, 80 Bq/m3 and 120 Bq/m3 determined in a national radon survey 1990-1991 (random sampling with 3 074 participants completing the study out of 5 000 invited). Assuming the non-measured houses to be statistically similar to those measured within each square, we may expect that in 6.4% of the detached houses the concentration will exceed the action limit 400 Bq/m3 (5.0% according to the national radon survey). The effect of foundation type and other building characteristics on the measurement activity and radon concentration is also discussed.
The on-going work to analyse the factors affecting the data base representativeness is expected to significantly broaden the usability of the data base for radon risk analysis, promoting the utilisation of data much larger than could be attained by random sampling studies.